We consider the performance analysis of generalized selection combining (GSC) for BPSK signals in time-correlated Rayleigh fading channels. For the coherent detection of BPSK signals, we employ pilot symbol aided modulation (PSAM) to provide channel side information (CSI) and the optimum GSC follows the maximal signal to noise ratio (SNR) criterion. To analyze the bit-error-rate (BER) of GSC for BPSK signals, we propose a novel method based on the Monte-Carlo integration technique to circumvent the high computational complexity occurred in the evaluation of analytical BER expression.

The number of users that can be supported by frequency-hopped, spread-spectrum multiple-access systems can be increased greatly by using multiuser demodulation and iterative decoding. In the receiver employed hard-decision multiuser demodulation followed by iterative decoding, users exchange decoded information with each other. Additional information from multiuser demodulation in the first decoding iteration is limited by the hard-decision output of the multiuser demodulator. The error-correction used was an errors-and-erasures Reed-Solomon (RS) decoder. We revisit hard-decision demodulation and conventional RS decoding. Hard-decision multiuser demodulation is modified to provide a soft output, which is then given to a nonbinary block turbo code with shortened RS codes as the constituent codes. An iterative multiuser decoding algorithm is developed to do soft multiuser interference cancellation. This soft receiver with soft demodulation and decoding is shown to be more resistant to multiuser interference and channel noise, especially at lower values of signal-to-noise ratio. The results show a great improvement in the ability of the system to support more users (more than three times in some cases), as compared with systems that erase all hits or employ hard-decision multiuser demodulation followed by RS code. We examine the proposed method for synchronous as well as asynchronous frequency-hopped systems in both AWGN and fading channels.

An iterative algorithm for joint decoding and channel estimation in frequency-hopping (FH) networks is proposed. In the proposed algorithm, soft decoder outputs are used in the iterative estimation of the time-varying variance of the additive interference resulting from the sum of the thermal noise, partial-band noise jamming, and other-user interference. The soft outputs are also used in the estimation of the independent random carrier phases and multiplicative Rayleigh fading coefficients in different frequency dwells. The estimation process is further enhanced through the insertion of known symbols in the transmitted data stream. The proposed iterative symbol-aided demodulation scheme is compared with the coherent scenario, where the channel state information is assumed to be known a priori at the receiver, for both convolutionally coded and turbo coded FH systems. The proposed iterative channel estimation approach is suited for slow FH systems where the channel dynamics are much slower than the hopping rate. This observation motivates the consideration of another robust approach for generating the log-likelihood ratios for fast hopping systems in additive white Gaussian noise channels. Simulation results that demonstrate the excellent performance of the proposed algorithms in various scenarios are also presented.

Joint channel estimation and decoding in a time-varying Rayleigh fading channel is considered. Knowing that the optimal solution or even the truncated near-optimal solution using iterative processing has an exponential complexity which hinders the practicability, a reduced complexity approach is proposed. This approach keeps the existing channel estimation and decoding schemes almost intact, while applying iterative processing to effectively exchange information between them. Thus, the complexity is rendered linear, and estimator adaptability can be easily established. We apply this approach to pilot symbol-assisted modulation (PSAM) and differentially modulated systems. It turns out that the performance is improved and the robustness to fading parameters is enhanced. Through simulations, we also show that the proposed method performs almost as well as the near-optimal design.

We consider the design of an array processor for space-time coded multi-antenna systems. As an alternative to the previously proposed zero-forcing method, in this paper, the maximum signal-to-noise ratio (SNR) criterion is used to obtain a balance between interference suppression and noise enhancement. Although the same in concept, this work differs from the conventional minimum mean-squared error method in that there is more than one desired signal dimension each corresponding to one of the space-time coded streams. It is shown that the number of linear filters required by the maximum SNR array processor is no more than the dimension of the signal space or the number of collaborating transmit antennas. The advantages of this design are highly improved performance and reduced decoding complexity.

The iterative MMSE multiuser detection paradigm is applied to the suppression of cochannel interference in the coded narrowband (multicell) uplink. The equivalent of multiple chips per bit (necessary for MMSE multiuser demodulation) is generated via multisensor reception, the array responses serving as user signatures. This receiver's robustness to overloading allows its sensor count to be much lower than the typical number of other-cell cochannel interferers. A permutation transmit diversity technique that exploits channel time-selectivity is proposed in order to enhance the interuser separation afforded by multisensor reception.

Power control is important in maintaining the communication link quality under fading and interference situations. In recent years, research on power control has been conducted toward reducing power consumption while maintaining reliable link quality. These previous works, however, did not try to optimize power control algorithms under practical constraints. In this paper, we first observe and provide some insights into the previous works on different aspects of power control schemes. These observations motivate us to propose a new power control scheme which has several novel features: it uses an adaptive optimal quantizer at the receiver for transmitting discrete feedback information and an adaptive quantization scaler/restorer followed by a loop filter at the transmitter. Optimal quantization minimizes the feedback information loss and the additional power control error caused by it, while the loop filter can be designed to achieve the lowest power control error. Optimization of the loop filter requires some computational power. Fortunately, filter self-design capability can be implemented. In that case, closed-loop power control can be seen as an instance of a general channel identification problem. Intuitive explanation, together with analysis and simulation results of the proposed scheme will be given.

We consider two iterative multiuser detection (MUD) schemes at the base station using code division multiple access (CDMA) signaling over a Rician fading channel. These two iterative MUD both employ soft interference cancellation (SIC) to eliminate the multiple access interference (MAI) from other users. After SIC, two filtering methods, one is based on matched filtering (MF) and the other one is based on the criterion of minimum mean-squared error (MMSE), are utilized to suppress residual interference. The forward error correction codes considered in this paper are convolutional codes and the BCJR decoding algorithm is used for acquiring the reliability of coded symbols. Since channel state information (CSI) can not be known perfectly at the receiver in practical situations, the effect of channel estimation error on system performance needs to be investigated. We modify the pilot aided demodulation (PAD) channel estimation scheme by taking advantage of iterative processing so that the accuracy of channel estimates can be improved as the number of iterations grows. Numerical results show the performance of the proposed MUD can achieve the single user bound with perfect CSI. The penalty when the CSI is estimated from the iterative PAD is about 2-3 dB which depends on the channel loads.

A code division access and switching method is applied to multibeam satellite fixed service communications. The access scheme is based on a spectrally efficient CDMA (SE-CDMA) while the onboard routing is based on code division switching (CDS). In this article we present the system architecture of the SE-CDMA and CDS, and evaluate the bit error rate performance of the end-to-end connection link. The CDS is a non-blocking switch with complexity linearly proportional to its size which can route calls from an uplink to a downlink beam without onboard demodulation or channel decoding. The SE-CDMA can achieve very low bit error rates at low signal to noise ratio, while it allows frequency reuse one (i.e., reuse of the available spectrum in every beam).

A soft iterative multisensor array receiver for coded multiuser wideband code-division multiple-access wireless uplink channels is proposed, such channels are typically both frequency- and time-selective. A new equivalent discrete-time synchronous representation is used to model asynchronous multiuser dispersive channels that employ, in general, random spreading. The proposed scheme suppresses multiuser interference over a wide range of user signal powers, by iteratively exchanging soft information between a minimum mean-square-error (MMSE) multiuser demodulator and a bank of single-user decoders, feeding back the outputs of the latter to aid in soft multiuser multisensor MMSE-RAKE demodulation and subtractive interference cancellation, in the second and subsequent iterations. It displays near-far resistance since it behaves like a successive interference subtracter across iterations. The array responses are obtained via a multipass estimation scheme that uses both (temporal) pilot symbols and soft estimates fed back from the decoders to effectively provide multisymbol pilot signals and thence successively refined estimates with increasing iterations, while seeking neither to rely on the array geometry nor to estimate the directions from which users' signals arrive at the receiver. Simulation studies indicate that this scheme performs close to the single-user case with a two-sensor receiver array, and perfect channel state information, after four iterations; alternatively, it allows significantly increased user capacity compared with conventional receivers, and suffers only a modest loss with estimated array responses.

Demodulation and decoding for frequency-hopped spread-spectrum multiple-access (FH/SSMA) systems have been traditionally conducted by conventional single-user (noncollaborative) demodulation and error- and erasure-correcting decoding techniques. In this paper, we study the demodulation and decoding aspects of collaborative multiuser reception for FH/SSMA and propose methods which increase the number of users the system can support. In particular, we propose and analyze the optimum maximum a priori probability demodulation of multiple symbols or type, and the use of iterative multiuser decoding after the demodulation. Since hits from one or two other users are the most likely hit events in FH/SSMA, the joint demodulation of two or of three users is performed based on likelihood ratio tests. M-ary frequency-shift keying modulation with noncoherent demodulation and Reed-Solomon codes with hard-decision minimum distance decoding are used in the FH/SSMA system. Results are derived for both synchronous and asynchronous frequency-hop systems. The performance of the proposed multiuser detector in additive white Gaussian noise and flat Rayleigh fading channels is evaluated. Scenarios when all simultaneous users or only a subset of them are collaboratively demodulated and decoded are simulated.

In attempting to find a spectrally and power efficient channel code which is able to exploit maximum diversity from a wireless channel whenever available, we investigate the possibility of constructing a full antenna diversity space-time turbo code. As a result, both three-antenna and two-antenna (punctured) constructions are shown to be possible and very easy to find. To check the decodability and performance of the proposed codes, we derive non-binary soft-decoding algorithms. The performance of these codes are then simulated and compared with two existing space-time convolutional codes (one has minimum worst-case symbol-error probability; the other has maximal minimum free distance) having similar decoding complexity. As the simulation results show, the proposed space-time turbo codes give similar or slightly better performance than the convolutional codes under extremely slow fading. When fading is fast, the better distance spectra of the turbo codes help seize the temporal diversity. Thus, the performance advantage of the turbo codes becomes evident. In particular, 10/sup -5/ bit-error rate and 10/sup -3/ frame-error rate can be achieved at less than 6-dB E/sub b//N/sub 0/ with 1 b/s/Hz and binary phase-shift keying modulation. The practical issue of obtaining the critical channel state information (CSI) is also considered by applying an iteratively filtered pilot symbol-assisted modulation technique. The penalty when the CSI is not given a priori is about 2-3 dB.

Use of multiple antennas at the transmitter and the receiver to increase the information capacity has received considerable attention. To exploit the capability of such systems in a practical way, it is of interest to ask if there are good codes with reasonable decoding complexity. Another question of interest is the simplicity of design. This issue is particularly important in the case where the number of antennas in transmitter and receiver are large and a space-time code may be to complex to design and decode. We illustrate a class of codes which are simpler to design and the interference across antennas is suppressed by linear processing at the receiver. The performance of such codes is excellent and much better than the group interference suppression techniques employed in conjunction with the space-time codes. Use of error detection schemes is shown to further improve the performance without much change in the design of codes and decoding complexity.

A new scheme termed orthogonal code division multiplexing (OCDM) is proposed for digital subscriber line (DSL) applications. Through the use of serial to parallel conversion and orthogonal codes, a "spreading gain" is achieved against narrowband interference without expanding the required transmission bandwidth of the information signal. Groups of orthogonal codes with different spreading gains and bandlimited (chip) filtering are used together with M-PAM data modulation. Spectral matching (to input power density that achieves the capacity of the copper cable channel) is performed followed by throughput optimization under a symbol error rate (SER) constraint in order to select the system parameters. The OCDM/DSL system performance is evaluated for several scenarios of interest.

Spectrally efficient CDMA (SE-CDMA) is a multiple access and modulation scheme proposed for multibeam geostationary communications satellites that provides fixed service directly to the end users. The SE-CDMA is designed to meet the service requirements for high spectral efficiency and low bit error rate (BER) given the on-board power limitations. Its implementation consists of a concatenated Reed-Solomon/turbo channel encoder, of a two step spreading operation isolating the satellite beams and users within each beam and of a M-ary PSK modulator. The up and down SE-CDMA links are considered to be coupled at the satellite by a CDMA routing circuit which performs a despreading and respreading operation. This paper presents the system architecture and the BER performance of the proposed SE-CDMA. Performance results show that the system can provide high spectral efficiency at very low BER and at low E/sub b//N/sub 0/.

We model and analyze the performance of both symbol-aided coherent demodulation (SAD) and pilot-aided coherent demodulation (PAD) schemes for multipath fading channels. An iterative algorithm for joint channel estimation and decoding is proposed. Both schemes are promising alternatives to differentially coherent demodulation for scenarios characterized with uncertainties in the carrier phase that make coherent demodulation infeasible. Multipath fading, the Doppler phenomenon due to user mobility and/or satellite movement, and different temperature and ventilation conditions at the sites of the various local oscillators that generate the transmitted signals cause the carrier phase uncertainty. Coherent demodulation requires the extraction of a reliable phase reference from the received signal. The traditional solution to these problems is differential demodulation that uses the phase of the previous symbol as reference but requires almost 3 dB (for M-ary PSK modulation in AWGN channels, it is more than that for fast fading channels) of additional signal-to-noise ratio (E/sup b//N/sub 0/) in order to achieve the same bit error rate (BER) as coherent demodulation. This problem is more severe in DS/CDMA systems which are limited by other-user interference. We propose two coded systems that use either PAD or SAD schemes for the reverse link. In both schemes, the estimate of the multipath channel is extracted by filtering either the known symbols in the SAD scheme or the pilot in the PAD scheme. The optimum pilot power in PAD and known symbols insertion rate in SAD are obtained.

Nonlinear amplifiers, present in both the base stations of cellular networks and transponders of satellite communications systems, are known to be one of the most undesired hardware distortions in the future DS/CDMA PCS systems. To mitigate the effects of nonlinearities, this paper proposes a design which modifies the input value of the correlator by employing a discriminator. We derive the optimal function of the discriminator. By computer simulation, we know we can improve the performance using the discriminator.

It has been shown that the capacity of a slow frequency hopped (SFH) spread spectrum multiple access (SSMA) system can be increased by using the joint demodulation and iterative decoding by a hard decision demodulator with Reed Solomon code as a forward error correcting code. In this paper, we revisited the FSK demodulation and the conventional RS decoding. The FSK demodulation was modified to do joint demodulation and provide the soft decision of the type in the presence of multi-user interference. The conventional RS decoding was replaced by a block turbo code consisting of two shortened RS codes with interleaving between them. An iterative multi-user decoding algorithm is developed to do a soft collaborative decoding. This 'soft core' of the decoder is shown to be more resistant to the multi-user interference and the channel noise. The results show a great improvement in the capacity of the system as compared to systems which erase the hits or which employ hard decision in the joint demodulation.

We consider the design of array processor for space-time coded multi-antenna systems. While the zero-forcing method was employed by Tarokh, Naguib, Seshadri and Calderbank (see IEEE Transactions on Information Theory, vol.45, no.4, p.1121-28, 1999), in this paper we seek to obtain a balance between interference suppression and noise enhancement. Although same in concept, this work differs from the conventional minimum mean square error (MMSE) method in that there are more than one desired signal dimensions each corresponding to one of the space-time coded streams. In this case, minimizing the mean squared error (MSE) involves averaging over data modulation and may complicate the presentation. Instead, maximizing the signal-to-noise ratio (SNR) can be handled more concisely. It will be shown that the number of linear filters required by the maximum SNR array processor is no more than the dimension of the signal space or the number of collaborating transmit antennas. The advantages of this design are highly improved performance and reduced decoding complexity.

The performance of the satellite communication system is limited by cochannel interference and multipath fading. In the present work, an iterative MMSE interference suppression in encoded TDMA satellite systems is proposed. Diversity of received signals at different hubs is assumed available and used to enable decoding of cochannel signals. The decoder outputs are fed back to help the cochannel interference suppression by employing a MMSE-base scheme. This process iterates as the decoder outputs converge to the transmitted symbols. A similar idea is also used to estimate channel response.

Orthogonal frequency division multiplexing (OFDM) is a promising multiplexing technique for high data rate transmission over wireless radio channels. In this paper, we consider a low mobility, wireless OFDM cellular system. Since a cellular system uses a frequency reuse concept to enhance the efficiency of spectral utilization, it introduces so called cochannel interference (CCI) which is one of the major sources of performance degradation. To mitigate the effect of CCI while satisfying required quality of service, the power level of each subchannel of the OFDM system should be allocated to and controlled at an optimal value such that the system throughput can be maximized by employing adaptive modulation. Due to the effect of the CCI in the multi-cell OFDM system, the optimal values of power levels are very difficult to find. A new distributed suboptimal power allocation algorithm for the multi-cell OFDM system is proposed to increase the system throughput. Numerical results show the advantages of the proposed algorithm over those of the equal power allocation algorithm in various channel conditions.

Multiuser detection for frequency hop spread spectrum multiple access (FH/SSMA) systems has been traditionally done by conventional single user decoding or by erasure techniques. We study the demodulation and decoding aspects of multiuser reception for SFH/SSMA and propose methods which improve the capacity of the system. In particular, we propose and analyze the joint demodulation of multiple symbols and the use of iterative decoding after the demodulation. For the hard decision demodulation, need Solomon codes of different users collaborate with each other after the joint demodulation to reduce the multi-user interference. This approach is then extended for the soft decision demodulation and decoding where a block turbo code consisting of two shortened RS codes is employed. Results are shown both for synchronous and asynchronous frequency hop systems and indicate considerable improvement over the conventional systems.

The theory of space-time codes for multiple antennas at the transmitter and receiver provided a useful framework for minimizing the probability of pair-wise frame error when the channel is unknown at the receiver. The design criterion for the codes were developed by averaging over all the channel states. In practice, a feedback channel of low bandwidth is usually available where the transmitter can get some channel state information from the receiver. The transmitter can use this information to change its power level or modulation or both. Since the channel state is a matrix instead of a scalar in the single antenna case, we first find out what is the sufficient statistic for space-time orthogonal block codes. It is shown that for a given channel state, the decoder essentially sees an AWGN channel with the signal to noise ratio scaled by the sufficient statistic. We discuss the optimal power control and adaptive modulation strategies. The effect of delay in passing the channel information from the receiver to the transmitter is examined. The results show a significant improvement (more than 10 dB in some cases) over the conventional system without feedback.

A new iterative receiver for joint detection and decoding of code division multiple access (CDMA) signals is presented. The new scheme is based on a combination of the minimum mean square error (MMSE) criterion and the turbo processing principle by Hagenauer (see Proc. Int. Symp. Turbo Codes and Related Topics, Brest, France, p.1-9, 1997). The complexity of the new scheme is of polynomial order in the number of users. The new scheme is applicable to two situations: (a) when the receiver is capable of decoding the signals from all users and (b) when the receiver is only capable of decoding the signals from a subset of users. In the first scenario, we establish that the proposed receiver achieves superior performance to the iterative soft interference cancellation technique under certain conditions. On the other hand, in the second scenario, we argue that the proposed receiver outperforms both the iterative soft interference canceler and the iterative maximum a posteriori (MAP) receiver because of its superior near-far resistance. For operation over fading channels, the estimation of the complex fading parameters for all users becomes an important ingredient in any multiuser detector. In our scheme, the soft information provided by the decoders is used to enhance this estimation process. Two iterative soft-input channel estimation algorithms are presented: the first is based on the MMSE criterion, and the second is a lower-complexity approximation of the first. The proposed multiuser detection algorithm(s) are suitable for both terrestrial and satellite applications of CDMA.

This paper introduces a code-division methodology into switching applications. The proposed method is applied in satellite-switched code-division multiple-access (SS/CDMA) systems for routing CDMA traffic channels on board the multibeam satellites. We present code-division switch (CDS) architectures, analyze the CDS performance, and assess its complexity. The CDS has been shown to route CDMA user channels without introducing interference. The proposed CDS architecture is nonblocking, and its hardware complexity and speed are proportional to the size of the switch. We also examine the amplitude distribution of the combined signal in the CDS bus and the interference evaluation of the end-to-end link in the proposed applications. Then we consider the problem of switch control under an optimum or a random algorithm and compare its complexity with the equivalent problem in time-multiplexed switching methods.

The performance of inhibit sense multiple access (ISMA) in a wireless CDMA environment for short burst data service is analyzed.. Data users are allowed to transmit short messages (up to several frames at a time) in a contention manner using the reverse common channels. In this paper, we assume that one such common traffic channel is allocated for this data service. Data users contend for the traffic channel using either the ISMA or slotted Aloha protocol. Markovian analysis is given and the throughput as well as the delay are computed assuming a fixed number of voice users in the system. Numerical results show that for this short data message scenario ISMA has a higher throughput than slotted Aloha.

Joint channel estimation and decoding in a time-varying Rayleigh fading channel is considered. Knowing that the optimal design has an exponential complexity which hinders the practicability, a reduced complexity approach is proposed. This approach keeps the existing channel estimation and decoding schemes almost intact, while applying the turbo or iterative processing principle to enable information exchange between them. Thus the complexity is rendered linear; and the filter adaptability can be easily established. We apply this approach to two commonly used modulation schemes for noncoherent detection, namely, pilot symbol assisted modulation (PSAM) and differential modulation, with the intention to overcome certain disadvantages seen in their demodulation. It turns out that the performance is improved and the robustness is enhanced. In this paper, we first give the system descriptions of the two applications. Their performances are then simulated and compared with the existing systems. To reveal the tradeoffs in practical applications, the two new designs are also compared with each other in terms of throughput, complexity, performance, and robustness.

This work presents a soft iterative minimum mean square error (MMSE) multiuser interference suppressor for coded direct sequence code division multiple access (CDMA) uplink channels that employ space-time processing at both the mobile transmitters and the base station receiver. With a new equivalent discrete-time synchronous model for randomly spread asynchronous multiuser CDMA channels, we develop an iterative scheme that exchanges soft information between an MMSE multiuser demodulator (that also concurrently performs subtractive interference cancellation) and a bank of single-user decoders. It performs well both at low and high received signal powers, and displays the successive cancellation property across iterations, exhibiting good near-far resistance. Introducing multisensor arrays at the receiver relaxes the limitations imposed by the inherent suboptimality of MMSE multiuser demodulation, in so far as it enhances interuser separation, now in the spatial sense, and thereby allows significant overload capability. Further, this work explores new territory by studying multisensor transmission in multiuser channels. A new multisensor block modulation scheme called space-time spreading, for mobile transmitters is proposed; it is observed to provide significant benefits especially when the spatial separation afforded by the use of multisensor receiver arrays, is otherwise limited.

This paper presents throughput evaluation of a satellite-switched code-division multiple-access (SS/CDMA) system which operates under demand assignment control. SS/CDMA provides both multiple access and switching to a geostationary multibeam satellite. Multiple access is resolved by space, frequency, and code division. Space division is introduced by the multibeam antennas that provide frequency reuse in each beam. The spectrum is then channelized into frequency bands where each band is accessed by code division for both the uplink and downlink. The satellite on-board performs the switching function, which is also based on compatible code-multiplexed switching. The switch may route both circuit calls and data packets which are assigned upon request. The on-board code-division switch operates under the control of a channel assignment algorithm. We provide channel assignment algorithms for optimum, suboptimum, and random switch operation. The system throughput has been evaluated for each case and compared. Performance analysis has been carried out for the case of the optimum switch scheduling. The analysis is based on a discrete-time Markovian model, and provides the call-blocking probabilities and data packet delays. Computer simulations have been used to evaluate the performance of the optimum, suboptimum, and random cases. It is shown that the circuit call-blocking probabilities achieved for these cases are almost the same. The optimum algorithm achieves the minimum data packet delay, while the performance of the suboptimum algorithm is slightly better than the random one. Furthermore, data packets may be routed via the switch with limited delays, even with a heavy load of circuit calls.

We consider the application of multicarrier modulation in a wireless cellular network in order to enable high-data rate communication and alleviate the multipath induced intersymbol interference (ISI). In this scenario, power control becomes crucial in enhancing the spectral and power efficiency. A conventional approach of maintaining the same link quality for all the subchannels, in other words, disregarding any possible post-demodulation processing, is considered first. This approach appears to have increasing power consumption as the number of subchannels increases. It also deteriorates the power control stability and convergence properties in a multicell network. We attribute this phenomenon to lack of frequency diversity exploitation, and thus, we propose to use channel coding and soft decoding as vehicles to profit from the (frequency) diversity advantage in addition to the coding advantage. Based on the soft decoding performance bound, a power allocation and control algorithm is proposed. It is shown through simulations that the proposed algorithm improves the power efficiency as the number of subchannels increases. It also provides a better convergence property and is able to "detect" and eliminate ill-conditioned subchannels. The advantages of using multicarrier modulation are thus reassured. Besides these enhancements, the proposed algorithm is simple and feasible in that it consists of only the traditional closed-loop power control algorithm and a target signal-to-interference ratio (SIR) reassignment at the receiver. Detailed channel information feedback from receiver to transmitter is not required.

In this paper, we present a blindly adaptive beamforming algorithm which is based on the second order interference estimation to maximize the received SINR. Using the desired signature code and the code orthogonal to it, a new code filter is introduced to decompose the received signals into two parts: desired information and interference. The above method motivates us to develop the QR-decomposition based dominant eigenmode search (QRD-DMS) algorithm which is more numerically stable than the DMS one. The corresponding wavefront systolic architecture is also proposed for the VLSI implementation. Compared to the families of minimum mean square error (MMSE) algorithms which need training sequences, we have completed the maximum SINR families by proposing the QRD-DMS method which not only blindly updates the beamforming weights but converges as fast as the QRD-RLS method.

Time reuse capture access (TRCA) is a packet access protocol proposed for personal communication services (PCS). The protocol is based on time reuse rather than frequency reuse. A time frame is divided into slots where each slot corresponds to one cell in the reuse cluster. Each mobile user in the cell may transmit in any time slot of the frame with probabilities associated with the cell it currently belongs. The protocol also exploits the power capture phenomenon. Simultaneous transmissions between users within a cell or in adjacent cells may result in packet capture which can increase the system throughput. This paper presents throughput and delay analyses when the time reuse factor is three. The system throughput (delay) is then maximized (minimized) by optimizing the values of the transmission probabilities for each time slot in the frame. A throughput improvement of 35% is achievable compared with a TDMA system.

We examine the applicability of symbol-aided demodulation (SAD) to Rayleigh and Rician fading channels, which are the channels most frequently encountered in terrestrial mobile communications and satellite communications, respectively. The effects of filtering and interpolation on the bit error rate (BER) performance of the SAD system are evaluated. The BER comparisons for optimum Wiener filtering and linear interpolation are carried out. The BER of SAD systems in the presence of Rician fading are compared with those of the fully coherent and differentially coherent schemes. It is established that SAD performance is close to that of perfectly coherent demodulation and significantly better than differentially coherent demodulation for a wide range of system and channel parameters.

The use of orthogonal CDMA for uplink satellite communication requires that all user codes arrive at the satellite despreader at a common time. To this end schemes for network synchronization among all users are necessary; these involve two stages: code acquisition and code tracking. We propose a code tracking loop that extends between the satellite and the ground terminal. The traditional code tracking loop is broken into two segments: the satellite segment of the loop conducts the phase difference extraction; while the ground segment of the loop at the Earth station performs the filtering and phase adjustment. The major difference of this tracking system and the traditional loops is the long propagation delay entering the loop delay and secondarily the satellite motion. The optimization of the tracking loop filter is undertaken with respect to the steady state error and the variance of the timing error. The performance of the code tracking loop scheme is evaluated for a single-beam fully-loaded SATCOM system.

We present a frequency hopping multiple-access concept suitable for multi-cell network architectures. A set of orthogonal frequency hopping (FH) patterns is assigned to the users in each cell and cells are differentiated by concatenating the user FH pattern with a shuffling sequence of the same hop rate. The use of guard times enables synchronous operation for the range of hopping rates of interest. The capacities of this network for the above FH/SS multiple access scheme with either coherent BPSK modulation or non-coherent BFSK modulation are evaluated for the AWGN channel. The effect of power control errors on the system performance is accounted for in each case. Similar results for asynchronous DS/SS cellular network (reverse link) are also reported for comparison purpose. The comparison is then extended to the Rayleigh fading channel. It is argued that synchronous FH/SSMA is implementable and can provide higher capacity than asynchronous DS/CDMA under certain conditions.

We investigate the application of joint decoding and channel estimation for DS/CDMA transmission over slow Rayleigh fading channels. Some known symbols are inserted in the encoded data stream to enhance the channel estimation process. An iterative algorithm that uses the decoding information, in addition to the information contained in the known symbols, to improve the channel parameters estimate is proposed. A is shown that the optimum scheme has a complexity which grows exponentially with the channel estimation filter length. Hence, we propose some alternative sub-optimum schemes with polynomial complexity. The different techniques that provide a tradeoff between simplicity of implementation and BER performance are compared.

Power control is important in maintaining the communication link quality under fading and interference situations. In previous years, research on power control has been conducted towards reducing the power consumption while maintaining reliable link quality. These previous works, however, did not try to optimize the power control algorithms under practical constraints. In this paper, we first provide a critical review of the previous works on different aspects of power control schemes. We then propose a new power control scheme which has several novel features: it uses an adaptive optimal quantizer at the receiver for transmitting discrete feedback information and an adaptive quantization scaler/restorer followed by a loop filter at the transmitter. Intuitive explanation of the effectiveness of the proposed design is provided. Analysis and simulation results are also given.

This paper provides an analytical framework for evaluating the performance of multi-cell orthogonal frequency division multiplexing (OFDM) networks in the presence of cochannel interference (CCI), frequency selective Rayleigh fading and additive white Gaussian noise (AWGN). Coherent and differential coherent modulation methods along with the imperfect power control mechanisms are considered in our analysis and the corresponding formulas of symbol error rate (SER) are derived. Moreover, forward error control coding schemes using bandwidth efficient inner trellis codes and outer Reed-Solomon codes are employed to improve system performance. The expressions of the asymptotic bit error rate (BER) of the inner trellis code and the SER of the concatenated code are analyzed. Finally, numerical results verify the effectiveness of the concatenated coding scheme in combating the channel fading and cochannel interference in OFDM networks.

This paper describes a medium access protocol for wireless ATM systems using time division multiple access with frequency division duplex. Voice and data mobiles are supported. In order to utilize the resource efficiently, information slots are allocated to mobiles that are in talkspurt only. When a mobile changes its state from silent to talkspurt, it has to request for an information slot. Real-time service is required for voice mobiles and so queueing is not allowed. The medium access scheme for voice mobiles is based on the non-collision packet reservation multiple access (NC-PRMA) which will ensure prompt delivery of voice packets. Data mobiles are considered as services that are delay-insensitive. Available bit rate services can be provided for (delay-insensitive) data mobiles. The scheme for data mobiles is based on the distributed queueing request update multiple access (DQRUMA). Numerical results are obtained through simulations and the performance of the proposed hybrid protocol is compared with that of PRMA and reservation TDMA. The results obtained show that the proposed protocol performs better in the sense of having smaller voice and data dropping probabilities under similar system conditions. Additional advantages are that optimization of the system throughput is easier and that the performance of voice mobiles is unaffected by instability in the data portion of the protocol.

An analysis of time jitter for codes used in synchronous CDMA systems is presented. The system uses a set of either quadratic-residue orthogonal codes or Gold quasi-orthogonal codes. Other user interference power is calculated for both codes taking into consideration the effect of time jitter. The effect of bandlimited chips is also addressed. A modified QR code design which is less sensitive to time jitter is proposed.

This paper evaluates the effects of nonlinearities on the performance of synchronous M-PSK code-division multiple-access (CDMA) systems for applications in multiple tiers. The system performance in terms of bit error rate (BER) and system capacity is discussed. The effects of utilizing different families of spreading code sequences such as orthogonal (OG), preferred-phase Gold (PG), and pseudonoise (PN) codes are explored. The desired operating point of a nonlinear amplifier is also determined in order to achieve the best performance.

We analyse the performance of multimedia virtual circuit switched networks. Voice, video and data traffic are considered. Although closed-form expressions for the end-to-end performance measures (probabilities of voice and video blocking, normalized voice processing loads, probabilities of video rate dropping and data queueing) are available, they cannot be applied directly because of the prohibitive computation required. As a result, fast and accurate approximation schemes based on the reduced load method are developed. The bandwidth allocated to a video call is varied; together with the step size of the virtual paths, the performance of the network can be controlled and optimized. Extensive numerical results are given to validate the approximation schemes developed.

We investigate the effect of limiting the access bandwidth of a video source. In some applications, the access bandwidth is less than the peak rate of the source and so dropping of packets occurs. A model is needed for this truncated source. The model suggested in the paper is based on the discrete-state continuous time Markov model proposed by Maglaris et al. (1988). We derive the mean and the autocorrelation function of the rate of this new model and show that these statistical properties agree with the experimental values.

In this paper, we provide: (a) An efficient method for simplifying MMPP source models, which can be used to solve multi-media queueing problems with complicated aggregate input traffic. The effectiveness of the method is verified through numerical examples. (b) An approximation algorithm for analyzing multi-link queueing networks with MMPP traffic sources and arbitrary configurations. In this algorithm, a queueing network is analyzed by decomposing it into individual queues, each with an arrival rate appropriately reduced by the packet dropping occurred in other queues. Each queue is then solved individually by using the simplified source models of (a) while the interdependencies with the other queues are accounted for through this reduced-load approximation. Any desirable end-to-end performance measure characterizing the quality of service (QOS) of the multi-media traffic can be calculated in this manner. The results obtained by the algorithm have been compared with simulation results, and the observed relative error is satisfactory.

The paper examines protocols for voice/data integration in hybrid packet radio networks consisting of satellite and terrestrial components. The same protocols can be useful in the context of mixed-media packet radio networks employing HF ground radio links and EHF SATCOM radio links. As part of the integration protocol the ground subnetwork employs code-division multiple-access with movable boundary in the code domain to serve both the voice traffic and the retransmitted data traffic; on the satellite subnetwork framed ALOHA with movable boundary is used for both data and voice. A complete analysis of the channel-access protocols for both traffic types is provided based on a multi-dimensional Markovian model of the integrated voice/data system. The performance of the hybrid network is evaluated in terms of voice and data throughput, voice blocking probability, and data delay. It is observed that appropriate splitting of the retransmission traffic between the satellite and ground subnets may increase the overall data throughput when the voice load is high.>

Multiple-access schemes are introduced and analyzed for the integration of voice and data traffic in packet radio networks using code-division multiple-access (CDMA). The multiple-access capability of the CDMA channel is used to accommodate several voice calls simultaneously, while the data users follow the ALOHA protocol with retransmission control and contend for the remaining (if any) multiple-access capability of that channel. The retransmission probabilities of the backlogged data users are

Circuit-switched networks of video users with different grades of service are considered. The authors consider admission control schemes that result in an increase to network throughput at the expense of some fraction of the information (rate) transmitted by the video sources being dropped. The end-to-end probability of blocking video requests and the associated probability of rate dropping of video sources are evaluated along each network path for the above control schemes and for an arbitrary circuit-switched network architecture. Instead of the exact expressions, which require prohibitive computational effort, the stochastic knapsack approximation and the Monte Carlo summation method are employed.>

Considers a network of low-earth-orbit (LEG) satellites with integrated voice and data traffic. The uplinks and downlinks of all satellites use code-division multiple-access (CDMA) with direct-sequence spread-spectrum (DS/SS) signaling. The satellites act as bent-pipes; no on-board processing or intersatellite links are present. Dual satellite diversity is used to mitigate the effects of shadowing. The authors find an optimal CDMA code allocation policy by minimizing a cost function consisting of the weighted sum of the blocking rates of voice and data traffic subject to performance requirements (bit error rates) of the two traffic types. The relative priority of voice and data can be adjusted via a weighting factor. The activity factor of the voice traffic (talkspurts and silent periods) is taken into account to increase bandwidth efficiency. The code allocation policy has the form of a movable voice/data boundary in the CDMA code domain.

We provide methods to evaluate the probabilities P(l, m-l|K), l=0, 1, ..., m and m/spl les/K of exactly l correct packet receptions in a group of m receivers, given that K packets are transmitted simultaneously from users employing direct-sequence spread spectrum (DS/SS) signalling schemes. This quantity is useful for the design and performance evaluation of protocols for admission control and dynamic code allocation in multiple-access spread spectrum packet radio networks intended for terrestrial or satellite applications. The evaluations are carried out for DS/SS networks employing BPSK modulation with coherent demodulation and convolutional codes with Viterbi decoding. Systems with geographically dispersed receivers and systems with colocated receivers are considered. Approximations based on the independent receiver operation assumption (IROA) and the Gaussian multivariate distribution are developed, and their accuracy is checked against the exact expressions derived for synchronous systems. The joint first error event approximation (JFEEA) is also developed for coded systems and compared to the IROA.>

The authors derive optimal admission policies for integrated voice and data traffic in packet radio networks employing code division multiple access (CDMA) with direct-sequence spread spectrum (DS/SS) signaling. The network performance is measured in terms of the average blocking probability of voice calls and the average delay and packet loss probability of data messages. The admission scheme determines the number of newly arrived voice users that are accepted in the network so that the long-term blocking probability of voice calls is minimized. In addition, new data arrivals are rejected if the mean delay or the packet loss probability of data exceeds a desirable prespecified level. A semi-Markov decision process (SMDP) is used to model the system operation. Then, a value iteration algorithm is used to derive the optimal admission control. Two models for the other-user interference of the CDMA system are considered: one based on thresholds and another based on the graceful degradation of the CDMA system performance, and their performance is compared. These admission policies find application in emerging commercial CDMA packet radio networks including cellular networks, personal communication networks, and networks of LEO satellites for global communications.>

Establishes that the compressive receiver is a practical interceptor of high performance. Given a signal of a particular duration, a compressive receiver can estimate simultaneously all frequency components within a set wide band. This processing is similar to a parallel bank of narrowband filters, which is the optimal detector of frequency-hopped signals. Furthermore, hop frequency is estimated to yield performance equal to the parallel filter configuration. The authors assume interference to be stationary, colored Gaussian noise, and present a model of the compressive receiver that contains all its salient features. Low energy coherence detection is achieved by taking the compressive receiver output as an observation and applying likelihood ratio theory at small signal-to-noise ratios. For small signals, this approach guarantees the largest probability of correct detection for a given probability of false alarm, and thus provides a reference, to which simplified or ad hoc schemes can be compared. Since the low energy coherence detector has an unwieldy structure, a simplified suboptimal detector structure is developed that consists of a simple filter, followed by a sampler and a square-envelope detector. Several candidates for the filter's response are presented. The performance of the low energy coherence detector based on compressive receiver observations is compared to the optimal filter-bank detector based on direct observations, thus showing the exact loss incurred when a compressive receiver is used. The performance of various simplified schemes, based on compressive receiver observations, is analyzed.>

Formulates and solves maximin and minimax detection problems for signals with power constraints. These problems arise whenever it is necessary to distinguish between a genuine signal and a spurious one designed by an adversary with the principal goal of deceiving the detector. The spurious (or deceptive) signal is subject to certain constraints, such as limited power, which preclude it from replicating the genuine signal exactly. The detection problem is formulated as a zero-sum game involving two players: the detector designer and the signal designer. The payoff is the probability of error of the detector, which the detector designer tries to minimize and the deceptive signal designer to maximize. For this detection game, saddle point solutions-whenever possible-or otherwise maximin and minimax solutions are derived under three distinct constraints on the deceptive signal power; these distinct constraints involve lower bounds on (i) the signal amplitude, (ii) the time-averaged power, and (iii) the expected power. The cases of independent and identically distributed and correlated signals are considered.>

Detectors based on one-step memory nonlinearities are introduced. Problems of discrimination between two arbitrary stationary m-dependent or mixing sequences of observations and problems of detecting a weak signal in additive stationary m-dependent or mixing noise are considered in this context. For each problem, the nonlinearity is optimized for performance criteria such as the generalized signal-to-noise ratio (SNR) and the efficacy, and is obtained as the solution to an appropriate linear integral equation. The schemes considered can be made robust to statistical uncertainties determined by two-alternating capacity classes, for the second-order joint PDFs (probability distribution functions) of the observations, and by bounds on the correlation coefficients of time shifts of the observation sequence for the third- and fourth-order joint PDFs. Evaluation of the performance of the schemes via simulation reveals significant gains over that of detectors employing memoryless nonlinearities or the independent and identically distributed nonlinearity.>

The problem of distributed estimation of a weak nonrandom location parameter theta in additive stationary dependent noise is addressed. Multisensor configurations with and without a coordinator are considered. Dependence in the sensor observations is described by m-dependent, phi -mixing, and p-mixing models. Two cases of interest are addressed: one in which sensor observations are dependent across time but independent across sensors and one in which sensor observations are dependent across both time and sensors. Numerical results on the performance evaluation of the various estimation schemes derived are presented and the relative performances of the various schemes are compared.>

Exact expressions for the probabilities P(l,m-l/k) of l correct packet receptions and m-l erroneous ones, out of total k packets contending in a slot, are presented for the case of frequency-hopped spread-spectrum random-access slotted networks employing random frequency hopping patterns. These expressions are difficult to evaluate numerically for values of m>3. However, their numerical analysis indicates that under light traffic conditions these probability values are very close to the ones provided by the independent receiver operation assumption, under which the distribution of multireception obeys the binomial law.>

An adaptive multichannel radiometer designed to detect frequency-hopping (FH) signals in complex signal environments is presented. Each channel updates its hop threshold to reflect the current environment and excise any persistent hop activity inconsistent with an FH signal from subsequent processing. This strategy allows the receiver to discriminate FH signals from any random noise or interference activity with relatively small degradations as compared to operation in stationary additive white Gaussian noise. Two data collection schemes are considered for the proposed receiver, both of which attempt overall decisions using fixed-length blocks of data. In the first scheme, block detection, successive decisions are based on consecutive, nonoverlapping blocks of data, whereas in the second, block-sequential detection, decisions are made each time a new datum is collected. The block-sequential scheme is shown to offer greatly reduced average signal detection times and, thus, is the preferred approach.>

Different spread-spectrum signaling schemes in a cellular mobile radio network are compared in terms of throughput and packet error probability. Bounds on the bit and packet error probabilities are derived for data modulation schemes with binary phase shift keying with noncoherent demodulation. Reed-Solomon coding is employed for error-correction purposes. In all cases, the effect of varying interference power (according to some inverse power of distance) of the desired signal, of the interfering signals, and of Rayleigh nonselective channel fading is accurately taken into account. The throughput in the mobile-to-base transmission mode is evaluated for the above data modulation, demodulation, and forward-error-control coding schemes. The comparison shows that, under the varying interference power model, the frequency-hopped scheme performs best among all schemes with the same bandwidth. Power control mechanisms are required to improve the performance of direct-sequence systems.>

The performance of synchronous and asynchronous, binary and quaternary (with and without offset) direct-sequence spread-spectrum multiple-access (DS/SSMA) communication systems using random signature sequences and arbitrary chip waveforms is investigated. The average probability of error at the output of the correlation receiver is evaluated using a characteristic-function approach for these systems. Numerical results are presented that illustrate performance comparisons between systems using random and deterministic signature sequences, synchronous and asynchronous systems, systems with rectangular or sinewave chip waveforms, and binary and quaternary systems with the same data rates and bandwidth. In all cases, the accuracy of the Gaussian approximation is also examined.>

The probability of capture is evaluated for the situation in which several transmitters use the same spread-spectrum code for the attention of a single receiver. The first stage in the capture mechanism is that of the acquisition of capture. The authors are concerned with the second stage, that of retaining capture, in the presence of interference from contending users. The probability of retaining capture is computed via accurate approximations and upper bounds for direct-sequence, frequency-hopped, and hybrid spread-spectrum signaling formats and for different data modulation and demodulation schemes. The calculation of the overall probability of capture is carried out for spread-spectrum systems with and without forward-error control. The ability to reject primary multiple-access interference in spread-spectrum radio networks is examined by computing the maximum number of users that may contend for the same receiver, without causing the probability of capture to fall below some desirable level.>

A numerical study of the performance of hybrid direct-sequence slow-frequency-hopped spread-spectrum random-access schemes using Reed-Solomon forward-error control coding is performed. The performance measures of interest are the maximum numbers of simultaneous transmitters that can be tolerated in the vicinity of a receiver at a desirable level of the bit (or symbol) error probability, the normalized throughput, and the packet error probability. For the case where all simultaneous signals have the same received power levels, a critical comparison of the performance of direct-sequence, frequency-hopped, and hybrid spread-spectrum systems with the same overall bandwidth expansion is conducted with respect to the above performance measures, and ranges of parameters, for which each system outperforms the others, are identified. It is also established that hybrid spread-spectrum random-access schemes have considerably higher throughput than the uncoded nonspread-spectrum ones, for the desired range of values of the packet error probability.>

Discrimination is considered between two possible sources based on dependent observations of their output. The discrimination problem is modeled by means of a general binary hypothesis test, the main emphasis being on situations that cannot be modeled as signals in additive noise. The observations are modeled as stationary m-dependent or rho -mixing processes. The structure of the discriminator is such that the observations are passed through a memoryless nonlinearity summed up to form a test statistic, which is then compared to a threshold. Only fixed sample size tests are considered. Four different performance measures, which resemble the signal-to-noise ratios encountered in the signal in additive noise problems, are derived under different problem formulations. The optimal nonlinearities for each of the performance measures are derived as solutions to various integral equations. For three of the four performance measures the authors have successfully obtained robust nonlinearities for uncertainty in the marginal and the joint probability density functions of the observations. Computer simulation results that demonstrate the advantage of using these nonlinearities over the i.i.d. nonlinearity under the probability of error criterion are presented.>

The author characterizes and evaluates the effect of simultaneous multiple partial-band noise or tone jammers and other user interference on a single communication link employing frequency-hopped spread-spectrum (FH/SS) signaling, M-ary frequency-shift keying (FSK) modulation with noncoherent demodulation, and Reed-Solomon coding. For the symbol error probability of these systems, the author derives exact expressions in the absence of multiple-access interference and tight upper bounds in the presence of other-user interference. Although the analytical methods are valid for any number of multiple jammers, the numerical study is restricted to the cases of two and three-partial-band noise and tone jammers. For fixed values of the spectral densities of noise jammers, or the energies per symbol of tone jammers, the worst-case fraction of the band that each jammer should use in order to maximize the error probability of the FH/SS or FH/SSMA system is evaluated. For the range of the signal-to-jammer power ratios examined, multiple-noise or multiple-tone jammers appear to have no advantage over single-tone jammers of equivalent spectral density or energy per symbol, but achieve approximately the same worst-case performance by jamming smaller fractions of the band.>

Two forms of spread-spectrum signaling, namely direct-sequence and hybrid (direct-sequence/frequency-hopped) are shown to provide high broadcast capability, especially when used in conjunction with forward-error-control coding schemes. The broadcast capability is defined as the maximum number of simultaneous distinct messages that can be transmitted to distant receivers from a single transmitter at a given bit-error rate. The quantity provides a useful measure of the capacity of hub-to-mobile or satellite-to-earth-station links of communication networks. When bursty data or voice traffic is dominant in such networks, the above forms of spread-spectrum CDMA provide a viable alternative to FDMA or TDMA. Ways of multiplexing the direct-sequence and hybrid signals are presented that use, respectively, distinct carriers, distinct pairs of orthogonal carriers, and only two orthogonal carriers for broadcasting the different messages. Systems with chip-synchronous signals and systems with random delays between the signals are considered. The average error probability of all systems is evaluated using the characteristic-function and Gaussian-approximation techniques. Besides the uncoded systems, systems using Reed-Solomon and convolutional codes are analyzed. A comparison of the broadcast capability of the different schemes is presented.>

The specific problem of detecting the presence of either binary-phase-shift-keyed (BPSK) signals or BPSK direct-sequence spread-spectrum (DS/SS) signals with a prefilter-delay-and-multiply (PFDM) device is considered. Using stationary process theory and Fourier analysis, the optimum PFDM structures for signal presence detection of BPSK signals with known bit rates and carrier frequencies and BPSK DS/SS signals with known chip rates and carriers in additive colored Gaussian noise are derived. The structures are optimum in the sense that they maximize the spectral signal-to-noise ratio (SNR) of an output periodic waveform which has fundamental frequency equal to the bit or chip rate of the signal. Two of the optimum structures that are derived and analyzed are the optimal prefilter-square device and the optimal PFDM with delay set to one half of the signal's bit or chip duration.>

A method for the evaluation of the probability of error of uncoded asynchronous frequency-hopped spread-spectrum multiple-access communications is presented. For systems with binary FSK modulation this method provides an accurate approximation and a tight upper bound to the bit error probability; for systems with M-ary FSK modulation, it provides tight upper bounds to the symbol error probability. The method enables the computationally efficient averaging of the error probability with respect to the delays, phase angles, and data streams of the different users. It relies on the integration of the product of the characteristic function of the envelope of the branch of the BFSK demodulator, which carries the desired signal, and of the derivative of the characteristic function of the envelope of the other branch. For sufficient frequency separation between the BFSK tones, the method can achieve any desirable accuracy. Moreover, the computational effort required for its evaluation grows linearly with the number of interfering users. In the M-ary case, tight upper bounds based on the union bound and the results of the binary case are derived. The method allows the effect of unequal power levels on other-user interference in FH/SSMA systems to be quantified accurately for the first time. The results indicate that the FH/SSMA systems suffer from the near-far problem, although less seriously than direct-sequence SSMA systems.>

The problem of designing matched filters that are robust against uncertainty in the statistics of a noise process modeled by two alternating Choquet capacities is addressed. The robust design is based on the maximum signal-to-noise ratio. The problem is formulated and solved for both discrete-time and continuous-time matched filters with uncertainty in either the autocorrelation function or the spectral measure of the noise. Explicit solutions that are characterized by the Hyber-Strassen derivative of the capacity generating the class with respect to a Lebesgue-like measure on a suitable interval are obtained.>

Minimax robust data fusion schemes for multisensor detection systems with discrete-time observations characterized by statistical uncertainty are developed and analyzed. Block, sequential, and serial fusion rules are considered. The performance measures used, and made robust with respect to the uncertainties, include the error probabilities of the hypothesis testing problem in the block fusion case and the error probabilities and expected numbers of samples or sensors in the sequential and serial fusion cases. For different sensor observation statistics, the minimax robust fusion rules are derived for two asymptotic cases of interest: when the number of sensors is large and when the number of times the fusion center collects the local (sensor) decisions is large. Moreover, for the case of identical sensor observation statistics and a large number of sensors, it is shown that there is no loss in optimality, if local tests using likelihood ratios and equal thresholds are used in the sequential fusion rule. In all situations, the robust decision rules at the sensors and the fusion center are shown to make use of likelihood ratios and thresholds that depend on the least-favorable probability distributions of the uncertainty class describing the statistics of sensor observations.>

The authors characterize multiple-access interference for cellular mobile networks, in which users are assumed to be Poisson-distributed in the plane and use frequency-hopped spread-spectrum signaling with a transmitter-oriented assignment of frequency-hopping patterns. Exact expressions for the bit error probabilities are derived for binary coherently demodulated systems without coding. Approximations for the packet-error probability are derived for coherent and noncoherent systems and these approximations are applied when forward-error-control coding is used. In all cases, the effects of varying interference power are accurately taken into account according to some propagation law. Numerical results are given in terms of bit-error probability for the exact case and throughput for the approximate analyses. Comparisons are made with previously derived bounds, and it is shown that these tend to be very pessimistic.>

The channelized receiver, which is optimal for the detection of unknown noncoherent frequency-hopped waveforms, bases its decisions on a fixed-length block of input data. A sequential method of interception is presented according to which whenever a new data element is collected, a decision is made as to the presence or nonpresence of a frequency-hopped waveform. If that decision is indeterminate, another data element is collected. An optimal sequential test is derived, under the assumption that the waveform signal-to-noise ratio (SNR) is known. It is shown that this sequential test requires less data, on average, than the fixed-length method to make a decision with the same reliability. A truncated sequential test is also derived where a decision is forced, if still indeterminate, after some fixed amount of data is collected. The truncated test is shown to improve the number of samples needed for a decision when the input SNR differs greatly from that assumed in the derivation of the test. Furthermore, it is shown that the truncated test yields a limited degree of robustness when the input SNR differs from that assumed.>

In this paper we address the problem of combatting combined interference in spread-spectrum communication links. We consider frequency-hopped spread-spectrum systems withM-ary FSK modulation and noncoherent demodulation which employ forward-error-control coding. The interference consists of partial-band noise jamming, nonselective Rician fading, other-user interference, and thermal noise. The coding schemes which we analyze include: ReedSolomon codes (with or without diversity and error-only, erasure-only, or parallel erasure/error decoding), binary, nonbinary, and dual-kconvolutional codes with and without side information (information about the state of the channel), and concatenated schemes (Reed-Solomon outer codes with either inner detection-only block codes or inner convolutional codes). In all cases we derive 1) the minimum signal-to-jammer energy ratio required to guarantee a desirable bit error rate as a function of ρ, the fraction of the band which is jammed, when the number of interfering users is fixed; and 2) the maximum number of users that can be supported by the system as a function of ρ, when the signal-to-jammer energy ratio is fixed.

The problem of designing optical receivers that are robust against uncertainty in the statistics of the observation process in photodetection is investigated. In particular, a modification in the design of the postdetection matched filter is proposed to account for possible uncertainty in the rate function of the incident light, the rate of the dark current, and in the statistics of the additive noise present at the input to the optical receiver. This design is based on a game-theoretic approach in which a filter is sought that has the maximum worst case output signal-to-noise ratio possible over the class of allowable statistics; that is, the design criterion is maximin signal-to-noise ratio. A general characterization of maximin robust matched filters for observed Poisson processes is presented in this context, and specific solutions for several useful uncertainty models are obtained. Numerical results are presented for a specific example to illustrate the performance of the proposed technique.

An accurate approximation is obtained for the average probability of error in an asynchronous binary direct-sequence spreadspectrum multiple-access communications system operating over nonselective and frequency-selective Rician fading channels. The approximation is based on the integration of the characteristic function of the multiple-access interference which now consists of specular and scatter components. For nonselective fading, the amount of computation required to evaluate this approximation grows linearly with the productKN, whereKis the number of simultaneous transmitters andNis the number of chips per bit. For frequency-selective fading, the computational effort grows linearly with the product KN2. The resulting probability of error is also compared with an approximation based on the signal-to-noise ratio. Numerical results are presented for specific chip waveforms and signature sequences.

The performance of noncoherent reception in synchronous and asynchronous hybrid direct-sequence/slow-frequency-hopped spread-spectrum multiple-access communication systems operating through additive white Gaussian noise channels is investigated. Systems with binary andM-ary frequency-shift-keying modulation and noncoherent demodulation, as well as systems with differential-phase-shift keying modulation and differentially coherent demodulation, are examined and their probability of error is evaluated for random frequencyhopping patterns and signature sequences. The multiple-access capability of noncoherent hybrid spread spectrum is shown to be superior to that of noncoherent purely frequency-hopped spread spectrum and inferior to that of noncoherent purely directsequence spread spectrum for systems with the same bandwidth expansion. Comparison of hybrid systems with coherent and noncoherent demodulation shows a considerable loss in the performance of the noncoherent systems.

The performance of noncoherent reception in direct-sequence spread-spectrum communications over specular multipath fading channels is investigated. Analytical and numerical results on the average probability of error are presented for direct-sequence spread-spectrum systems employing binary orM-ary frequency-shift-keying modulation with noncoherent demodulation and differential phase-shift-keying modulation with differentially coherent demodulation.

The problem of minimax robust coding for classes of multiple-access channels with uncertainty in their statistical description is addressed. We consider1)discrete memoryless multiple-access channels with uncertainty in the probability transition matrices and2)discrete-time stationary additive Gaussian multiple-access channels with spectral uncertainty. The uncertainty is modeled using classes determined by two-alternating Choquet capacities. Both block codes and tree codes are considered. A robust maximum-likelihood decoding rule is derived which guarantees that, for ali two-user channels in the uncertainty class and all pairs of code rates in a critical rate region, the average probability of decoding error for the ensemble of pairs of random block codes and the ensemble of pairs of random tree codes converges to zero exponentially with increasing block length or constraint length, respectively. The channel capacity and cutoff rate regions of the class are then evaluated.

The performance of noncoherent reception in direct-sequence spread-spectrum multiple-access communications systems is investigated for additive white Gaussian noise channels. Analytical and numerical results on the probability of error are presented for binary andM-ary frequency-shift-keying data modulation with noncoherent demodulation and differential-phase-shift-keying data modulation with differentiallycoherent demodulation. Both synchronous and asynchronous systems are analyzed. Systems which employ deterministic as well as random signature sequences are considered. The multiple access capability of noncoherent DS/SS systems is evaluated and compared to that of coherent DS/SS systems with the same parameters. The comparison shows that the loss in the performance of DS/SSMA systems due to noncoherent reception can be considerably larger than the loss incurred to noncoberent single-user systems operating in additive Gaussian noise.

The performance of synchronous and asynchronous hybrid direct-sequence/slow-frequency-hopped spread-spectrum multiple-access communications over additive white Gaussian noise channels is examined. Systems employing binary or quaternary phase-shift-keying modulation with coherent demodulation are investigated. Both deterministic and random signature sequences and frequency-hopping patterns are considered and several possible assignments for them are discussed. It is shown that the multiple-access capability of hybrid spread-spectrum is superior to that of pure frequency-hopped spread-spectrum, and inferior to that of pure direct-sequence spread-spectrum for systems with identical bandwidth expansion which employ the same data modulation and demodulation scheme and random hopping patterns and signature sequences.

The performance of coherent direct-sequence spread-spectrum communications over specular multipath fading channels is investigated. The average probability of error of the correlation receiver is derived for an arbitrary number of paths with deterministic or random gain coefficients. The gain coefficients, delays, and phase angles of any two distinct paths are modeled as mutually independent random variables. Numerical results for several values of the system and channel parameters are presented.

Decision designs that are insensitive to modeling uncertainty are developed for the Chernoff bounds on the performance of binary hypothesis testing problems. These designs are based on observations with statistical uncertainty modeled by using general classes generated by 2-alternating capacities. The results are illustrated for the two cases of independent identically distributed observations with uncertainty in the probability distribution and discrete-time stationary Gaussian observations with spectral uncertainty, and they are applicable to several other cases as well. For the Chernoff upper bounds on the error probabilities, a "robust" decision design based on the I/kel/hood-ratio test between a least-favorable pair of probability distributions or spectral measures, respectively, is derived. It is then shown that for all elements in the uncertainty class this choice of likelihood ratio guarantees the exponential convergence of the aforementioned Chernoff bounds to zero as the number of observations or the length of the observation interval increases.

The problem of minimax robust coding for classes of channels with uncertainty in their statistical description is addressed. Specific consideration is given to: 1) discrete memoryless channels with uncertainty in the probability transition matrices; 2) discrete-time stationary Gaussian channels with spectral uncertainty; and to uncertainty with classes determined by 2-alternating Choquet capacities. Both block codes and convolutional codes are considered. A robust maximum-likelihood decoding rule is derived; the rule guarantees that, for all channels in the uncertainty class and all rates smaller than a critical rate, the average probability of decoding error for the ensemble of random block codes and the ensemble of random time-varying convolutional codes converges to zero exponentially with increasing block length or constraint length, respectively. The channel capacity and cut-off rate of the class are then evaluated.

The problem of robust design is considered in the context of testing hypotheses concerning the rate function of an observed point process. Designs that are insensitive to uncertainty in the rate functions are developed by applying a minimax formulation to two different measures of signal-to-noise ratio. Uncertainty in the rate is modeled by using general classes of rate measures generated by Choquet 2-alternating capacities, and solutions are characterized for this case by a Radon-Nikodym type derivative between such classes. It is shown that for uncertainty within capacity classes the robust decision design developed for the signal-to-noise ratio is also robust in a weaker sense for the Chernoff upper bounds on the error probabilities. Furthermore, the use of such a test guarantees the exponential convergence of these bounds to zero with increasing length of the observation interval for all rates in the uncertainty class.

Approximations are obtained for the average probability of error in an asynchronous direct-sequence spread-spectrum multiple-access communications system. Both binary and quaternary systems are considered, and the chip waveforms are allowed to be arbitrary time-limited waveforms with time duration equal to the inverse chip rate. The approximation is based on the integration of the characteristic function of the multiple-access interference. The amount of computation required to evaluate this approximation grows only linearly with the product of the number of simultaneous transmitters and the number of chips per bit. The accuracy of the approximation is extremely good in most cases, but it can be improved, if necessary, by an application of a series expansion. Numerical results are presented for specific chip waveforms and signature sequences.

Bounds and approximations are obtained for the average probability of error in an asynchronous slow-frequency-hopped spread-spectrum multiple-access communications system with non-coherent binary frequency-shift-keyed (FSK) data transmission. Both nonselective fading and wide-sense-stationary uncorrelated-scattering fading are considered.