This work considers a layered coding approach for efficient transmission of data over a wireless block fading channel without transmitter channel state information (CSI), which is connected to a limited capacity reliable link, known as the bottleneck channel. Two main approaches are considered, the first is an oblivious approach, where the sampled noisy observations are compressed and transmitted over the bottleneck channel without having any knowledge of the original information codebook. The second approach is a non-oblivious decode-forward (DF) relay where the sampled noisy data is decoded, and whatever is successfully decoded is reliably transmitted over the bottleneck channel. The bottleneck channel from relay to destination has a fixed capacity C. We examine also the case where the channel capacity can dynamically change due to variable loads on the backhaul link. The broadcast approach is analyzed for cases that only the relay knows the available capacity for next block, and for the case that neither source nor relay know the capacity per block, only its capacity distribution. Fortunately, it is possible to analytically describe in closed form expressions, the optimal continuous layering power distribution which maximizes the average achievable rate. Numerical results demonstrate the achievable broadcast rates.

This work considers the problem of transmitting two streams with different delay constraints over a single input single output slowly flat fading channel, where only the receiver has perfect channel state information (CSI). One stream has stringent delay constraints, while the other stream has a relaxed delay constraint. This is modeled by a block fading channel for the first stream, where a broadcast approach is employed. The second stream is encoded over a longer block and views an ergodic fading channel. Several encoding and decoding strategies are considered. In all cases the delay constrained stream is decoded in presence of the other stream, which can be decoded only after its long codeword is fully received. The non-delay constrained part is decoded after canceling out the first stream from the received signal. We derive here a closed form solution for the optimal layering power distribution, which maximizes the expected sum-rate, where the layering is employed on the delay constrained stream.

In a dynamic wireless network where a source terminal communicates with a destination, in presence of other users in the network, it is worthwhile to consider oblivious relaying strategies of a relay in close proximity to the source transmitter. The source-relay channel is assumed to be a fixed gain additive white Gaussian noise (AWGN) channel due to source-relay colocation. The source-destination and the relay-destination transmissions are subject to block flat fading dynamics. A perfect channel state information (CSI) at the respective receivers only is assumed. With the expected throughput as a performance measure, we incorporate a two-layer broadcast approach into a cooperative strategy based on the decode-and-forward scheme - Sequential Decode-and Forward (SDF). The broadcast approach splits the transmitted rate into superimposed layers, corresponding to a "bad" and a "good" channel states, allowing better adaptation to the actual channel conditions. The achievable rate expressions for the SDF strategy are derived under the broadcast approach for multiple settings including single user, multiple-input single-output (MISO) and the general relay setting using successive decoding technique, both numerically and analytically. Continuous broadcasting lower bounds are derived for the MISO and for the oblivious cooperation scenarios.

We consider the problem of multiuser detection for randomly spread direct-sequence (DS) code-division multiple access (CDMA) over flat fading channels. The analysis focuses on the case of many users, and large spreading sequences such that their ratio, which is the system load, is kept fixed. Spectral efficiency of practical linear detectors such as match-filter and decorrelator employing successive interference cancellation (SIC) at the receiver is derived. This is used to extend the notion of strongest users detectors for SIC receivers. The strongest users detectors system design relies on an outage approach where each user transmits in a single layer (fixed rate), and only users experiencing good channel conditions may be reliably decoded, while the other users are in outage, i.e., not decoded. In this scheme, iterative SIC decoding is studied, and it is shown that for equal power users, the optimal rate allocation, for maximizing the expected spectral efficiency, is equal rates for all users. This outage approach analysis is extended for multilayer coding broadcast approach per user. The expected sum-rate, under iterative decoding with linear multiuser detectors, is optimized, and the optimal layering power distribution is obtained. For small system loads, the achievable spectral efficiency with the continuous broadcast approach and a linear matched filter detector exhibits significant gains over the single layer coding approach.

A single-server queue concatenated with a multilayer channel encoder is considered. The main focus of this work is on minimization of the expected delay of a packet from entering the queue until completion of successful service. Tight bounds are derived for the expected delay for different numbers of coded layers. Numerical optimization is applied to find the optimal resource allocation minimizing the average delay. Similar bounds are also derived for the case of continuous layering. It is demonstrated that code layering may give pronounced performance gains in terms of delay, which are more impressive than those associated with throughput. This makes layering more attractive when communicating under stringent delay constraints.

We consider the problem of transmission of a common message from a single source to two cooperating destination users, exhibiting different signal-to-noise ratios (SNR). The cooperation is performed over a noiseless link with a total capacity limitCcoop, which the users may utilize for conversation. In order to maximize the achievable throughput, it is suggested to employ a multilayer broadcast approach combined with multisession cooperation, such that every session another layer can be decoded. Since the cooperation link is limited, an efficient cooperation scheme is required. The relaying user performs compression which utilizes the side information of the decoder in the Wyner-Ziv (WZ) spirit. In multisession cooperation the compression is in the form of successive-refinement WZ coding. After the first user decodes a layer, the minimal required information is sent back to the second user to allow both of them to decode the message. The minimal required information is achievable with random coding using a binning strategy. While it is expected that multisession cooperation will be more efficient than a single-session cooperation we show that the maximal throughput is achieved with a single session cooperation. This is shown for the nonfading channel and for the ergodic fading case. Nonetheless, for the block fading channel it is already known that multilayering and multisession outperforms single session cooperation.

We study the problem of a single source transmitting, over a non-fading Gaussian channel, to two cooperating users exhibiting different signal-to-noise ratios (SNR). Both users are required to decode the source message. This model is similar to the broadcast channel with a common message, however the users cannot reliably decode the message independently, and require some form of conversation in order to decode. The cooperation is performed over a noiseless link with a total capacity limit Ccoop, which the users may utilize for conversation. In order to maximize the achievable throughput, it is suggested to employ the multi-layer broadcast approach combined with multi-session cooperation, such that every session another layer can be decoded. Since the cooperation link is limited, an efficient cooperation scheme is required. The relaying user performs compression accounting for the side information of the decoder in the Wyner-Ziv (WZ) spirit. In multi-session cooperation the compression is in the form of successive-refinement WZ coding. After the first user decodes a layer, the minimal required information is sent back to the second user to allow both of them to decode the message. The minimal required information is achievable with random coding using a binning strategy. While it is expected that multi-session cooperation will potentially be more efficient than a single-session cooperation we show that the maximal throughput is achieved in a single session.

We consider the problem of transmitting a Gaussian source on a slowly fading Gaussian channel, subject to the mean-squared error distortion measure. The channel state information is known only at the receiver but not at the transmitter. The source is assumed to be encoded in a successive refinement (SR) manner, and then transmitted over the channel using the broadcast strategy. In order to minimize the expected distortion at the receiver, optimal power allocation is essential. We propose an efficient algorithm to compute the optimal solution in linear time , when the total number of possible discrete fading states. Moreover, we provide a derivation of the optimal power allocation when the fading state is a continuum, using the classical variational method. The proposed algorithm as well as the continuous solution is based on an alternative representation of the capacity region of the Gaussian broadcast channel.

rdquoWe investigate the performance of the broadcast approach for various fading distributions, which correspond to different models of partial transmit channel state information (CSI). The first model considered is the quantized limited feedback. In this model, the receiver can send as feedback only a finite number of bits describing the fading gain. We derive the optimal power allocation for the broadcast approach for the quantized feedback model. For a Rayleigh fading channel, numerical results here show that if the feedback word can be longer than one bit, the broadcasting gain becomes negligible, due to diminished channel uncertainty. The second partial transmit CSI model is a stochastic Gaussian model with mean and variance information, which is commonly used for modeling the channel estimation error. In a single-input single-output (SISO) channel, this model also corresponds to the Ricean fading distribution, for which we derive maximal achievable broadcasting rates. We further consider a multiple-input single-output (MISO) channel, and derive the optimal power allocation strategy in a broadcast approach. Numerical results here show that uniform power allocation is preferable over beamforming power allocation in the region where broadcasting gain over single level coding is non-negligible.

Conventional hybrid automatic retransmission request (HARQ) is usually used to maximize throughput. However, high throughput is achieved at the expense of high latency. We study a novel broadcasting HARQ strategy. The multi-layer broadcast approach is suitable for the case where transmitter has no channel state information (CSI), which is the case with HARQ schemes as well. The broadcast approach enables the receiver to decode rates, which are matched to every fading gain realization. That is, the higher the fading gain realization, the more layers are reliably decoded. The broadcast approach combined with HARQ enables achieving high throughput with low latency. In a broadcast HARQ scheme every code layer supports HARQ independently. Thus HARQ is applied in every transmission block to undecoded layers only, which highly increases the broadcast approach efficiency. In this paper, both broadcast chase combining (BCC) HARQ and broadcast incremental redundancy (BIR) HARQ are studied in the limit of infinitely many layers, and for finite level coding. Interestingly, with continuous broadcasting the BCC-HARQ is found to closely approximate the BIR-HARQ, while using a sub-optimal broadcasting power distribution.

We consider the problem of transmitting a Gaussian source on a slowly fading Gaussian channel, subject to the mean squared error distortion measure. The channel state information is known only at the receiver but not the transmitter. The source is assumed to be encoded in a successive refinement manner, and then transmitted over the channel using the broadcast strategy. In order to minimize the expected distortion at the receiver, optimal power allocation is essential. We propose an efficient algorithm to compute the optimal solution in linear time O(M). Moreover, we provide a derivation of the optimal power allocation when the fading state is a continuum, using the classical variational method. The proposed algorithm as well as the continuous solution is based on an alternative representation of the capacity region of the Gaussian broadcast channel.

This work considers the problem of communication between a remote single transmitter and a destined user, with helping colocated users, over an independent block Rayleigh-fading channel. The colocation nature of the users allows cooperation, which increases the overall achievable rate, from transmitter to destination. The transmitter is ignorant of the fading coefficients, while receivers have access to perfect channel state information (CSI). We propose, for this setting, a multilayer broadcast transmission approach. The broadcast approach enables enhanced cooperation between the colocated users. That is due to the nature of broadcasting, where the better the channel quality, the more layers that can reliably be decoded. The cooperation between the users is performed over additive white Gaussian noise (AWGN) channels, with a relaying power constraint, and unlimited bandwidth. Three commonly used cooperation techniques are studied: amplify-and-forward (AF), compress–and-forward (CF), and decode–and-forward (DF). These techniques are extended by using the broadcast approach for the case of relaxed decoding delay constraint. For this case, a separate processing of the layers, which includes multisession cooperation is shown to be beneficial. Further, closed-form expressions for infinitely many AF sessions and recursive expressions for the more complex CF are given. Numerical results for the various cooperation strategies demonstrate how the multisession cooperation outperforms conventional relaying techniques.

This paper introduces extensions for the broadcast approach for a multi-input multi-output (MIMO) block fading channel, with receiver only channel state information (CSI). Previous works have not been able to fully characterize the fundamental MIMO broadcasting upper bound. As it seems that analytical solution for this problem is quite difficult to achieve, we consider here sub-optimal schemes, for which achievable rates may be computed. In particular, finite level coding over a MIMO channel instead of continuous layering is analyzed, the expressions derived for the decoding probability regions allow numerical computation of finite level coding upper bounds. Noticing that the gains of two level coding over a MIMO channel are rather small, we consider sub-optimal techniques, which are more straightforward to implement. Among these techniques is the multiple-access channel (MAC) approach with single level coded streams, which is similar in concept to V-BLAST. Closed form expressions for probabilities of decoding regions here are derived, allowing numerical evaluation. We further consider multi-access permutation codes (MAPC). A Hadamard transform is compared with a suggested diagonal permutation code, which are shown to have similar performance, while diagonal permutation has lower implementation complexity. For all approaches, we derive information theoretic upper bounds of achievable rates.

This paper presents an ASIC hardware architecture of a multi-antenna access point (AP), compliant with 802.11a/b/g standard and designed for outdoor metropolitan-area deployment. The AP has a six-antenna array and its PHY layer implements single-user maximum ratio combining (MRC) and transmit beamforming (BF), as well as downlink multi-user fourfold spatial division multiple-access (SDMA). The clients use standard WiFi 802.11a/b/g networks interface cards (NICs).

A two-hop relay fading channel is considered, where only decoders possess perfect channel state information (CSI). Various relaying protocols and broadcasting strategies are studied. The main focus of this work is on simple relay transmission scheduling schemes. For decode-and-forward (DF) relaying, the simple relay cannot buffer multiple packets, nor can it reschedule retransmissions. This gives rise to consideration of other relaying techniques, such as amplify-and-forward (AF), where a maximal broadcasting achievable rate is analytically derived. A quantize-and-forward (QF) relay, coupled with a single-level code at the source, uses codebooks matched to the received signal power and performs optimal quantization. This is simplified by a hybrid amplify-QF (AQF) relay, which performs scaling, and single codebook quantization on the input. It is shown that the latter is optimal by means of throughput on the relay-destination link, while maintaining a lower coding complexity than the QF setting. A further extension of the AQF allows the relay to perform successive refinement, coupled with a matched multilevel code. Numerical results show that for high signal-to-noise ratios (SNRs), the broadcast approach over AF relay may achieve higher throughput gains than other relaying protocols that are numerically tractable

A broadcast transmission strategy for the slowly fading Gaussian multiple input multiple output (MIMO) channel is introduced. This broadcast strategy is an extension of earlier work of Shlomo Shamai (Shitz) (see IEEE Int. Symp. Inform. Theory - ISIT'97, p.150, 1997) for the single input single output (SISO) channel. It is assumed that perfect channel state information (CSI) is known at the receiver end only. This strategy facilitates adapting the reliably decoded rate to the actual channel state without having any feedback link to the transmitter. Transmission of layered coded information is motivated by the theory of majorization. We derive the basic equations characterizing achievable rates of the strategy. Several ad-hoc approximations to the achievable region are considered and their performance is compared with the SISO setting and the ergodic capacity. It has been demonstrated that a single layer outage approach is reasonably efficient in the MIMO setting in terms of the average reliably decoded rate.

A new suboptimal demodulator based on a singular value decomposition for estimation of unitary matrices is introduced. Noncoherent communication over the Rayleigh flat fading channel with multiple transmit and receive antennas, where no channel state information is available at the receiver is investigated. Codes achieving bit-error rate (BER) lower than 10/sup -4/ at bit energy over the noise spectral density ratio (E/sub b//N/sub 0/) of 1.6-1.9 dB from code restricted capacity limit were found. At higher data rates, computation of code restricted capacity is impractical. Therefore, the mutual information upper bound of the capacity attaining isotropically random unitary transmit matrices was used. The codes achieve BER lower than 10/sup -4/ at E/sub b//N/sub 0/ of 3.2-6 dB from this bound, with coding rates of 1.125-5.06 bits per channel use, and different modulation decoding complexities. The codes comprise a serial concatenation of turbo code and a unitary matrix differential modulation code. The receiver employs the high-performance coupled iterative decoding of the turbo code and the modulation code. Information theoretic arguments are harnessed to form guidelines for code design and to evaluate performance of the iterative decoder.

A broadcast transmission strategy for the slowly fading Gaussian multiple-input multiple-output (MIMO) channel is introduced. This broadcast strategy is an extension of the single-input single-output (SISO) broadcast approach. Perfect channel state information (CSI) is assumed known at the receiver end only. This strategy facilitates to adapt the reliably decoded rate to the actual channel state without having any feedback link to the transmitter. Transmission of layered coded information is motivated by the theory of majorization. We derive the basic equations characterizing achievable rates of the strategy. Several ad hoc approximations to the achievable region are considered and their performance is compared with the SISO setting and the ergodic capacity. It has been demonstrated that a single-layer outage approach is reasonably efficient in the MIMO setting in terms of the average reliably decoded rate. A multiple-access channel (MAC) broadcast approach is also applied for the MIMO case, and demonstrated to be relatively efficient.

A new suboptimal demodulator based on iterative decision feedback demodulation (DFD), and a singular value decomposition (SVD) for estimation of unitary matrices, is introduced. Noncoherent communication over the Rayleigh flat-fading channel with multiple transmit and receive antennas, where no channel state information (CSI) is available at the receiver is investigated. With four transmit antennas, codes achieving bit-error rate (BER) lower than 10/sup -4/ at bit energy over the noise spectral density ratio (E/sub b//N/sub o/) of -0.25 dB up to 3.5 dB, with coding rates of 1.6875 to 5.06 bits per channel use were found. The performance is compared to the mutual information upper bound of the capacity attaining isotropically random (IR) unitary transmit matrices. The codes achieve BER lower than 10/sup -4/ at E/sub b//N/sub o/ of 3.2 dB to 5.8 dB from this bound. System performance including the iterative DFD algorithm is compared to the one using Euclidean distance, as a reliability measure for demodulation . The DFD system presents a performance gain of up to 1.5 dB. Uncoded systems doing iterative DFD demodulation and idealized pilot sequence assisted modulation (PSAM) detection are compared. Iterative DFD introduces a gain of more than 1.2 dB. The coded system comprises a serial concatenation of turbo code and a unitary matrix differential modulation code. The receiver employs the high-performance coupled iterative decoding of the turbo code and the modulation code. Information-theoretic arguments are harnessed to form guidelines for code design and to evaluate performance of the iterative decoder.

A new sub-optimal demodulator based on a singular value decomposition (SVD) for estimation of unitary matrices is introduced for non-coherent communication over the Rayleigh flat fading channel with multiple transmit and receive antennas, where no channel state information (CSI) is available. Codes achieving bit error rate lower than 10/sup -4/ at bit energy over the noise spectral density ratio (E/sub b//N/sub 0/) of 1.7 dB from code restricted capacity limit or of 3.2 dB and higher from a mutual information upper bound of the capacity attaining isotropically random unitary transmit matrices, were found with coding rates of 1.125 to 5.06 bits per channel use, and different modulation decoding complexities. The codes comprise a serial concatenation of turbo code and a unitary matrix differential modulation code. The receiver employs the high performance joint iterative decoding of the turbo code and the modulation code.