Cognitive radio (CR) technologies have long been studied and continue to attract attention for their potential to enhance wireless spectrum sharing and utilization. In CR ad hoc wireless networks (CRAHWNs), unlicensed secondary nodes (SNs) are equipped with CR transceivers capable of continuously scanning for unoccupied wireless channels. This scanning process is managed through channel-hopping (CH) rendezvous schemes, which assign CH sequences to SNs, enabling dynamic control of frequency-hopping patterns used by their CR transceivers. Traditional CH schemes operate under “global” labeling, where all SNs share an identical mapping between logical channels in their CH sequences and the transmission/reception frequencies utilized by their CR transceivers. However, when SNs operate with differing channel-to-frequency mappings—arising from regional variations or restricted access to a common frequency set—rendezvous attempts fail, preventing data exchange. Despite its importance, the development of CH sequences capable of supporting “autonomous” labeling, enabling SNs with diverse channel-to-frequency mappings to achieve successful rendezvous, remains unexplored. This paper introduces a novel class of asynchronous “universal-label” CH sequences designed to seamlessly adapt to both global and autonomous labeling frameworks. Performance evaluations demonstrate that the proposed sequences achieve an optimal balance of essential properties. These advancements enable efficient spectrum sharing and utilization in CRAHWNs, even under challenging autonomous labeling scenarios.

In this article, two-hop transmissions in channel-hopping (CH) cognitive-radio wireless networks (CRWNs) with buffer-aided relays are studied. The two-hop relay (rendezvous) model is improved using symmetric CH sequences, which are more suitable for practical CRWNs than the old model using asymmetric ones. A proper buffer storage size in the relays is formulated to minimize the two-hop data-transfer latency. The lower and upper bounds of the steady-state maximum-time-to-rendezvous (MTTR) and throughput with relays are derived with formal proofs. Using the proper buffer size, both metrics can be better controlled and found no worse than twice the original MTTR and throughput without relays. To explore the impacts of utilization probability of Internet of Things (IoT) devices to their performances in a small-scale CRWN, the single-region rendezvous-success (RS) rate and throughput are formulated. For a large-scale CRWN, the model is extended to formulate the multiregion RS rate and throughput with and without relays for the first time. Numerical studies of these analytical models are performed and validated with computer simulations. In summary, the transmission range and number of rendezvousing IoT devices are improved in the large-scale CRWN with relays, and the multiregion RS rate and throughput with relays are found better than those without relays.

In this paper, two new families of asynchronous channel-hopping (CH) sequences—Symmetric Maximum-Length CH (SML-CH) Sequences and Relaxed Difference Set (RDS) of Maximum-Length CH (RDSML-CH) Sequences—are constructed. For the first time, the maximum-length sequences (also called as m-sequences) are used to create 2-D CH matrices with a reduced number of “jump” columns. The two new constructions carry the desirable properties of maximum rendezvous diversity (MRD) and even channel use (ECU) and, more importantly, have the shortest periods in their categories of constructions. Another contribution is on the embedment of optical orthogonal codes (OOCs) to eliminate “stay” columns in the RDSML-CH matrices, thus resulting in the first and only family of the asynchronous CH sequences that can asymptotically achieve the theoretical period lower bound. To support the RDSML-CH construction, a new shortened RDS algorithm for creating the OOCs with smaller weight-to-length ratios than the existing ones is proposed. Numerical and simulation analyses show that the two new constructions have a good balance of short period, MRD, ECU, short time-to-rendezvous (TTR) mean, small TTR variance, and short maximum-TTR, thus more suitable for practical CR wireless networks than the existing constructions.

Besides designing channel-hopping (CH) sequences, finding good methods of communication rendezvous is important for improving spectral efficiency and alleviating traffic load of cognitive radio (CR) networks. In this paper, new asynchronous-symmetric CH sequences are constructed and their (channel overlap, even-channel-use, and pairwise shift invariance) properties are investigated. Their operations in three transmission scenarios involving different basic collision-avoidance mechanisms are analyzed and compared in terms of maximum time-to-rendezvous, rendezvous-success (RS) rate, and RS variance. This paper shows that the new sequences have shorter period, larger cardinality, and smaller RS variations than other CH sequences, thus supporting greater number of secondary users and more frequent and uniform rendezvous in the CR networks.

Protocol sequences have recently been studied in collision channels without feedback for dynamic multiple-access mobile applications, such as wireless sensor and vehicular ad hoc networks. In this paper, a method of systematically adding mark chips into some families of prime sequences for the construction of protocol sequences with adjustable duty factors is investigated. Not relying on computer simulation only, new theoretical models of throughput and its variance are formulated to analyze the sequence performance by means of average hit probabilities. Numerical studies and computer simulations are performed to validate the new analytical models. With adjustable duty factors, these protocol sequences support flexible exchange between throughput and the number of active users.

In this paper, a novel code-shifting technique, which can be added atop asynchronous incoherent optical code-division multiple-access systems without complex modifications, is proposed and studied. The new technique divides the time-slot (or chips), in which the optical pulses of codewords are located, into g equal-width sub-chips; every pulse is randomly and independently shifted to start at one of the g sub-chips of its own chip, where g > 1 is a positive integer. The effects of this random g-shift to the cross-correlation property and performance of asynchronous optical codes (of periodic cross-correlation functions of at most one) are formulated. Although it is found that the periodic crosscorrelation functions are worsened to 3 - 2/g, our theoretical and computer-simulation results show an unconventional finding that code performance improves as g increases.

In this paper, two new families of "synchronous" optical codes, called 1-D and 2-D synchronous multilevel prime codes (SMPCs), are constructed and analyzed for incoherent optical code-division multiple access (O-CDMA). With the benefit of synchronous transmission, the new algebraic codes have expanded cardinality and substantially increase the number of possible subscribers, thus potentially supporting rapid code switching and obscure-nature (privacy and military) communications. In addition, the SMPCs can be partitioned into a tree structure of multiple levels of subsets of codewords. Different cross-correlation functions are obtained from codewords coming from different subsets; code orthogonality is achieved among the codewords within each subset of the lowest level. As code performance is a strong function of the cross-correlation function, the performance of the codewords in use can be adjusted by choosing the subsets from which the codewords originate. The performances of the SMPCs are formulated and also verified by computer simulation. Our analysis shows that code performance can be optimized by applying an "in-order" subset-by-subset code-assignment method to simultaneous users. Besides supporting time-spreading and wavelength-time O-CDMA schemes, the new codes also find applications in passive optical networks and in spectral-amplitude-coding schemes.

In this paper, the extended Reed-Solomon codes are modified to construct a new family of 2-D codes for asynchronous optical code-division multiple access (O-CDMA). In addition of having expanded and asymptotically optimal cardinality, these 2-D asynchronous optical codes can be partitioned into multiple tree structures of code subsets, in which code cardinality is a function of the (periodic) cross-correlation value assigned to the subset. The performance of these 2-D optical codes is analyzed and compared with that of the multilevel prime codes. Our results show that the unique partition property of the new optical codes supports a trade-off between code cardinality and performance for meeting different system requirements, such as user capacity and throughput. In addition, the multiple tree structures of the new codes potentially support applications that require rapid switching of many codewords, such as in O-CDMA-network gateway or in strategic environments where code obscurity is essential.

Due to power restriction of some laser sources, the use of same power (per bit duration) in optical codes in incoherent optical code-division multiple-access multimedia networks with variable quality-of-services was previously studied. In those studies, the performance of double-weight codes with the maximum cross-correlation function λc= 1 was exactly analyzed, but an approximation was only applied to the codes with λc> 1 due to mathematical complexity. In this paper, an exact analytical model of double-weight codes with an arbitrary λc≥ 1 is formulated and validated with computer simulation. This general, more accurate performance model is also useful for power-sensitive applications, such as in-service monitoring and fiber-fault surveillance in optical networks and sensor-identification in fiber-sensor systems using optical codes for address or user identification.

Coded modulation with Reed-Solomon (RS) codes has recently been proposed to use atop of frequency-hopping code-division multiple access (FH-CDMA) in order to increase data transfer rate. In this paper, a new method on implementing the modulation and demodulation processes of the RS codes in RS/FH-CDMA is proposed. With the use of our specially "pruned" trellis and the Viterbi algorithm, both processes are simplified and so are their computational complexities. The performance and computational complexities of this RS/FH-CDMA system are analyzed and verified with computer simulation.

Due to power restriction of some laser sources, the use of same optical power in every bit duration (i.e., same bit power) of double-weight codes in optical code-division multiple access (O-CDMA) with differential quality-of-service (QoS) was previously studied and analyzed with approximated performance. In this paper, an exact analytical model is formulated and validated with computer simulation, establishing accurate relationships of code weight, chip power, and bit power with code performance. Besides better understandings the role of power control in variable-weight O-CDMA with differential QoS, such accurate analytical model is important for power-sensitive applications, such as in-service monitoring and fiber-fault surveillance in optical networks and sensor-identification in fiber-sensor systems with the use of optical codes.

To provide differential quality-of-service in optical code-division multiple access, variable-weight codes were proposed and analyzed under a conventional assumption of identical power in every optical pulses (i.e., same chip power). To account for power limit of some laser sources, we assume the use of same optical power in each bit duration (i.e., same bit power) in this paper. For illustration, we focus our study on double-weight codes under both same-chip-power and same-bit-power assumptions. A new and better approximate performance-analytical model for double-weight codes is developed and validated with computer simulation. We show that heavy-weight codes do not always perform better than light-weight codes under the same-bit-power assumption, contrary to previous findings under the same-chip-power assumption.

Two-dimensional (2D) wavelength-time coding schemes have recently been studied for optical code-division multiple access. In this paper, we propose a new family of 2D codes, which use "synchronized" quadratic congruence sequences for wavelength hopping and (n, w, 2, 2) optical orthogonal codes for time spreading. The new 2D codes provide larger (asymptotically optimal) code cardinality than other 2D codes by relaxing the maximum cross-correlation function from one to two. Our study shows that the heavier code weight supported by our 2D codes results in better code performance than some of previously proposed 2D codes (of the maximum cross-correlation functions of one and two). The new codes potentially support more subscribers and heavier bursty asynchronous traffic with gradual performance degradation as traffic load increases.

In this paper, we propose a "two-level" frequency-hopping code-division multiple-access (FH-CDMA) scheme for wireless communication systems. The new scheme provides flexibility in the selection of modulation codes and FH patterns. By partitioning the modulation codes, our two-level scheme can be modified to support more possible users without increasing the number of FH patterns. The performance and spectral efficiency (SE) of the scheme are analyzed. Our results show that the partitioned two-level FH-CDMA scheme supports higher data rate and greater SE than Goodman's frequency-shift-keying FH-CDMA scheme under some conditions.

In this paper, a two-level frequency-hopping code-division multiple-access (FH-CDMA) scheme, which is flexible in the selection of modulation codes and FH patterns, is studied. This two-level FH-CDMA scheme can be modified by the partition of the modulation codes in order to support more subscribers without increasing the number of FH patterns. The performance and spectral efficiency (SE) analyses show that our scheme supports higher data rate and SE than Goodman's frequency-shift-keying FH-CDMA scheme, making the former more attractive for wireless communication systems.

In this paper, a new family of "multilevel" wavelength-time codes with expanded code cardinality, flexible code length, zero autocorrelation sidelobes, and the maximum cross-correlation function of any arbitrary positive integer n is constructed and analyzed. Besides having asymptotically optimal cardinality, the new codes can be partitioned into n levels, such that the cross-correlation function between any two code matrices in level l is at most l, where 1 ¿ l ¿ n. The hardlimiting performance of these multilevel codes is analyzed algebraically. Our results show that the unique partitioning property of these multilevel codes supports a trade-off between code cardinality and performance for meeting different system requirements, such as user capacity and throughput.

In this paper, we study a new class of two-dimensional codes, here called multi-level prime codes, with expanded code cardinality by relaxing the maximum cross-correlation function to any arbitrary positive integer n. Besides having asymptotically optimal cardinality and zero autocorrelation sidelobes, these multi-level prime codes can be partitioned into n levels, such that the cross-correlation function between any two code matrices in level l is at most l, where lisin[1, n]. The performance of the new codes in an optical code-division multiple-access system with hard-limiting detection is analyzed. Our results show that the unique partition property of the multilevel prime codes allows the trade-off between code cardinality and performance for meeting different system requirements, such as capacity and throughput.

In this paper, a new family of two-dimensional (2D) wavelength-time codes, which uses one-dimensional (1D) optical orthogonal codes (OOCs) with the maximum cross-correlation value of two as the time-spreading codes is proposed and analyzed. By relaxing the cross-correlation functions to at most two, the new 2D codes can provide larger code cardinality to accommodate more subscribers. They also have better performance than 1D optical codes because they can support heavier code weight, for a given code length. Moreover, our numerical examples show that the heavier code weight supported by our new 2D codes results in better performance than some of our previously proposed 2D codes, which have the maximum cross- correlation value of one.

The traditional chip-synchronous assumption used in the analyses of optical codes in optical code-division multiple access gives a pessimistic performance upper bound, while a more realistic chip-asynchronous assumption gives a more accurate performance. It is also known that a hard-limiter can be placed at the front end of an optical decoder to reduce the effects of multiple-access interference and the near-far problem. In this paper, the "hard-limiting" performance of two-dimensional (2-D) optical codes is analyzed under the chip-asynchronous assumption. We apply a Markov-chain method for a more accurate analysis, which can be generalized to 2-D optical codes with arbitrary maximum cross-correlation values. The performance of 2-D optical codes with the hard-limiting and chip-asynchronous assumptions is also compared with the soft-limiting and chip- synchronous assumptions.

In this paper, we study a new family of carrier-hopping prime codes (CHPCs), so-called quadratic- congruence CHPC (QC-CHPC), with expanded code cardinality by relaxing the maximum cross-correlation function to two (i.e., lambdac= 2). An application of the new code is for multicode keying in wavelength-time optical code-division multiple access (O-CDMA), in which each user is assigned M = 2mcode matrices to represent m data bits per symbol. The advantages are that 1) a lower baud multicode-keying O-CDMA system can support a higher bit rate and 2) user's code obscurity is enhanced, as compared to the transmission of one code matrix (per user) for each data bit one in conventional "on-off keying" O-CDMA systems. Our performance analysis shows that there is a trade-off between the code performance and hardware complexity/speed related to the choice of M, for a given bit rate.

Quadratic-congruence carrier-hopping prime codes (QC-CHPCs) with zero autocorrelation sidelobes, cross-correlation values of at most two, and expanded cardinality were recently constructed for wavelength-time "multicode-keying" optical code division multiple access (O-CDMA) for improved data throughput and code obscurity. To support multimedia services with different discrete bit-rate requirements, "multiple-length" QC-CHPCs are constructed algebraically in this paper. In contrary to conventional single-length codes, our analysis shows that the performance of these multiple-length codes improves as the code length decreases, thus supporting services prioritization in O-CDMA. Moreover, the relationship of the normalized spectral efficiency (NSE) and code lengths of the multiple-length QC-CHPCs is studied. Our results show that the NSE improves as the number of simultaneous users with short code matrices increases, which, however, decreases the total number of simultaneous users in the system. The choice of which code-length distribution to use depends on whether system efficiency or total number of simultaneous users is more important.

In this paper, the performance of two-dimensional optical codes with cross-correlation functions of at most one, such as the carrier-hopping prime code (CHPC), is analyzed without the traditional chip-synchronous assumption. In addition to the Gaussian approximation method, we apply a combinatorial technique for a more accurate analysis, especially when there exists only a small number of simultaneous users. It is known that an optical hard-limiter can be placed at the front end of a receiver to reduce the effect of multiple-access interference. We also derive the hard-limiting error probability under the chip-asynchronous assumption. Finally, the performance improvement of the CHPC with the hard-limiting and chip-asynchronous assumptions is studied.

To support multimedia services with different discrete bit-rate requirements, families of multilength optical codes, such as the carrier-hopping prime code (CHPC), extended CHPC, and multiwavelength optical orthogonal code, were recently constructed for wavelength-time optical code-division multiple-access (O-CDMA). In this paper, the performance of these multilength optical codes in a multimedia O-CDMA system with a variable-weight operation are analyzed. Our study shows that short-length codes generate stronger interference than long-length codes. This supports services prioritization in O-CDMA. Our study also shows that code weight is a more important factor than code length in determining code performance (i.e., quality of service).

To support multimedia services with different discrete bit-rate requirements, families of multilength optical codes, such as the carrier-hopping prime code (CHPC), extended CHPC, and multiwavelength optical orthogonal code, were recently constructed for wavelength-time optical code-division multiple-access (O-CDMA). In this paper, the performances of these multilength optical codes in a multimedia O-CDMA system with a variable-weight operation are analyzed. Our study shows that short-length codes generate stronger interference than long-length codes. This supports services prioritization in O-CDMA. Our study also shows that code weight is a more important factor than code length in determining code performance.

In this paper, we propose the use of Reed-Solomon (RS) code sequences for modulation on top of a frequency-hopping code-division multiple access (FH-CDMA). Our analysis shows that this RS/FH-CDMA scheme supports higher data rate but worse performance than the conventional -ary frequency-shift-keying FH-CDMA scheme. These two schemes provide a tradeoff between the data rate and the performance.

In this paper, a multiple-rate, multicarrier direct- sequence code-division multiple-access (MC/DS-CDMA) system with the use of two-dimensional orthogonal variable spreading factor (2-D OVSF) codes is studied. They are the first 2-D codes that can be generated by the tree structure commonly used by one-dimensional OVSF codes in wideband-CDMA. The 2-D OVSF codes also preserve orthogonality among code matrices with different spreading factors in the code tree. The performances of the proposed system with RAKE receivers employing equal-gain and maximal-ratio combining methods are analyzed and compared. Our results show that our multiple-rate MC/DS-CDMA system with the 2-D OVSF codes is more suitable for a non-fading additive white Gaussian noise channel or a Rician weak-fading channel. However, in a Rayleigh fading channel, the system performance gets worse because the orthogonality of the 2-D OVSF codes is destroyed by the strong fading effect.

In this paper, we propose a new fast frequency-hopping code-division multiple-access (FH-CDMA) scheme, in which prime sequences are used for modulation on top of FH-CDMA. The new prime/FH-CDMA scheme supports higher data rate than the conventional M-ary frequency-shift-keying FH-CDMA (MFSK/FH-CDMA) scheme, at the expense of worsening cross-correlation values. The performance and normalized spectral efficiency of both schemes are analyzed and compared. Our results show that the prime/FH-CDMA scheme, in general, utilizes similar system efficiency as the MFSK/FH-CDMA scheme and that both schemes provide a trade-off between data rate and performance.

We propose a modified bipolar-bipolar code for wavelength-time optical code-division multiple access. The new code uses Gold sequences for time spreading, in place of Barker sequences in our previously proposed bipolar-bipolar code. The new code has a cardinality of N+1 times greater than the original code, because there exist N+2 Gold sequences, but only one Barker sequence, for a given code length N. Since the new code is modified from the original code, there is no increase in hardware complexity. The average performance of the new code is analyzed and found very close to that of the original code. In other words, the new code can support greater cardinality without worsening average performance or increasing hardware complexity.

This paper first reviews an extended family of carrier-hopping prime codes (CHPC) with expanded (asymptotically optimal) cardinality and ideal correlation properties. They are designed for wavelength-time optical code-division multiple-access (O-CDMA) applications with broadband lasers, in which there are extra numbers of available wavelengths. The performance of the extended CHPC is analyzed and compared to a wavelength-grouping scheme using the original CHPC. Furthermore, the traditional pessimistic chip-synchronous assumption used in performance analysis of 2D codes with cross-correlation functions of at most one is removed for a more realistic error probability. In addition to the Gaussian approximation, we introduce in this paper a more accurate combinatorial method, especially when the number of simultaneous users is small.

A two-dimensional, so-called wavelength-hopping time-spreading, optical code-division multiple-access coding scheme has recently been studied for supporting more subscribers and simultaneous users than conventional one-dimensional techniques. To further improve the numbers of subscribers and simultaneous users without sacrificing performance, a new code, which utilizes the unipolar version of bipolar codes (for wavelength hopping) and optical unipolar codes (for time spreading), is studied and analyzed in this paper. A programmable and integratable coding hardware based on arrayed waveguide gratings is designed for this coding scheme.

A new family of two-dimensional (2-D) wavelength-hopping time-spreading codes, which employs wavelength hopping algebraically under prime-sequence permutations on top of time-spreading optical orthogonal codes, is studied and analyzed. Different from other 2-D codes, our new codes allow the number of wavelengths and code length to be chosen independently and, at the same time, the code cardinality is a quadratic function of the number of wavelengths without sacrificing the maximum cross-correlation value (i.e., still at most one). They are particularly suitable for high bit-rate optical code-division multiple-access systems with broadband mode-locked lasers, in which the number of time slots is very limited, and system capacity can only be grown by increasing the number of wavelengths, rather than code length. Finally, a novel wavelength-aware detector for wavelength-hopping time-spreading codes is discussed and shown to provide improved code performance.

In this paper, we study the use of shifted carrier-hopping prime codes (CHPC) for multicode keying in wavelength-time optical code-division multiple-access (O-CDMA) systems. By using the code's good properties of zero autocorrelation sidelobes and cross-correlation functions of at most one, each user is assigned M=2/sup m/ shifted copies of its own code matrix to represent m data bits per symbol with minimal interference. The advantages of our scheme are that: 1) a lower baud rate O-CDMA system can now support a higher bit rate; 2) no network synchronization is required; 3) no M-fold increase in code cardinality is needed, as compared with standard M-code keying; 4) the numbers of optical encoders and decoders can be reduced from M to one per user, a major hardware cost savings; and 5) user code confidentiality is enhanced, as compared with the transmission of one code matrix for each data-bit one in conventional "on-off keying" O-CDMA schemes. Finally, our performance analysis shows that there is a tradeoff between the error probability and the choice of M.

A new class of two-dimensional (2-D) carrier-hopping prime codes with very good correlation properties and significantly expanded cardinality was recently constructed for wavelength-hopping time-spreading optical code-division multiple access (CDMA). In this paper, the performance of the new codes is analyzed and compared to a carrier-grouping scheme of the original carrier-hopping prime codes. In addition, the traditional chip-synchronous assumption used in the performance analysis of optical CDMA is removed, for the first time, for 2-D optical codes with cross-correlation functions of at most one, showing improvement and better accuracy in error probabilities.

In this paper, a new class of carrier-hopping codes for frequency-hopping wireless and wavelength-hopping fiber optic code-division multiple-access (CDMA) is constructed by modifying the original carrier-hopping prime codes. The new carrier-hopping prime codes possess expanded and asymptotically optimal cardinality, zero autocorrelation side-lobes, and cross-correlation functions. The performance analysis of the new codes will be given and compared to the carrier-grouping scheme. Under the best scenario that a central controller is used to uniformly distribute all simultaneous users over groups of carriers in the grouping scheme, the results show that the new codes perform as good as the grouping scheme if the traffic load is heavy

In this paper, a new family of wavelength-time codes, called multiple-wavelength optical orthogonal codes (MWOOCs), is constructed. The new codes use prime sequences for wavelength permutations on top of a time-spreading OOC. The advantage is that the code cardinality is a quadratic function of the number of wavelengths. This property is particularly useful in high bit-rate optical CDMA applications, in which the number of time slots is very restricted. This limitation can be compensated and even improved in our MWOOCs by increasing the number of available wavelengths, through the use of lasers with broadened supercontinuum spectrum

Multiple-wavelength optical orthogonal codes (MWOOCs) with autocorrelation sidelobes and cross-correlation values of both at most one were recently proposed for wavelength-time optical code-division multiple-access (O-CDMA) systems. The codes have cardinality as a quadratic function of the number of wavelengths and find applications in high bit-rate O-CDMA systems with broadband supercontinuum lasers, in which the number of available wavelengths is larger than the number of time slots. To support multimedia services with different bit-rate and quality-of-service requirements, a new class of multiple-length constant-weight MWOOCs with autocorrelation sidelobes of zero and cross correlations of at most one is constructed algebraically in this paper. The performance of these new codes in an O-CDMA system with double-media services is analyzed. In contrary to conventional single-length codes, our study shows that the performance of these multiple-length codes improves as the code length decreases. This unique property supports "prioritization" in O-CDMA.

A two-dimensional wavelength-hopping time-spreading coding scheme has been recently studied for optical code-division multiple access, owing to maturity in fiber-Bragg grating (FBG) and arrayed-waveguide grating (AWG) technologies. Most of the optical codes designed for the scheme were based on an assumption, which may not be true in high data-rate systems, that the number of chips is higher than the number of available wavelengths. To provide a flexible code design without the assumption, this paper studies and analyzes a new family of carrier-hopping prime codes with expanded cardinality and ideal correlation properties. Flexible implementation of programmable AWG- and FBG-based coding hardware by using the wavelength-shift property of the codes is discussed.

This paper presents two new constructions of two-dimensional (2D) spreading codes, here designated as 2D orthogonal variable spreading factor (OVSF) codes, for multicarrier direct-sequence code-division multiple-access (MC/DS-CDMA) systems. Orthogonality of these spreading codes is employed to improve the bandwidth efficiency and interference rejection capability of the DS-CDMA systems. The 2D OVSF codes are constructed by generalizing one-dimensional (1D) OVSF codes and they possess ideal correlation properties. As illustrated in this paper, these 2D OVSF cods preserve the orthogonality of different spreading factors in an OVSF code tree, supporting multimedia services with a variety of data rates.

A family of multiple-length, constant-weight wavelength-time optical orthogonal codes is constructed. The codes are suitable for integration of multimedia services with different quality-of-service requirements and signaling rates in optical code-division multiple-access systems with fiber-Bragg-grating array encoders/decoders. The rate and service quality are dynamically matched to users' needs through the assignment of different-length wavelength-time code patterns, as illustrated.

To simultaneously support multimedia services with different signaling rates and quality-of-service requirements in optical code division multiple access (CDMA) networks, a new class of multilength, constant-weight optical orthogonal codes (OOCs) with good correlation properties is constructed algebraically in this paper. The performance of these new OOCs in an optical CDMA system with double-media services is analyzed. In contrast to conventional CDMA, our study shows that the performance of these multilength OOCs worsens as the code length increases, allowing prioritization in optical CDMA. Finally, an application of these multilength OOCs to integrate different types of multimedia services is briefly discussed.

Unequal-spaced channel-allocation techniques have been studied to reduce four-wave-mixing (FWM) crosstalk in high-capacity wavelength-division multiplexing (WDM) transmission systems. In this paper, we develop two algorithms that are helpful to system designers in quickly determining the total numbers of FWM crosstalk falling into the operating band and each channel of such WDM systems. By knowing the numbers, one can adjust the system parameters, such as the minimum channel spacing, in order to keep a good balance on the adverse effects of FWM crosstalk and interchannel interference or to avoid the assignment of channels at locations with the worse crosstalk.

Studies on multilevel frequency-shift keying/frequency-hopping spread-spectrum multiple-access (MFSK/FH-SSMA) wireless systems are mostly assumed to support only one type of media. Since a flexible support of multimedia services is expected in future, the use of conventional fixed-length FH patterns alone cannot serve these multi-rate, multi-media wireless systems. In this paper, a family of double-length, fixed-weight generalized prime hopping patterns, well suited for wireless systems with double-media services, is reviewed. The performances of the prime hopping patterns and conventional random hopping patterns in MFSK/FH-SSMA wireless systems are analyzed and compared for one type of media first and extended to two types. Our results show that the proposed scheme performs better than the conventional one with random hopping patterns.

A family of "multi-length", constant-weight frequency-hopping codes is constructed in this paper. The new codes are suitable for the integration of multimedia services (e.g., data, voice, image, and video) with different quality-of-service requirements and signaling rates in future telecommunication systems, where the rate and service quality are dynamically matched users' needs through the assignment of different-length frequency-hopping patterns. The use of the new codes in multi-rate, multimedia systems with frequency-hopping code-division multiple-access is illustrated.

We demonstrate experimental all-optical code-division multiplexing (AO-CDM) systems using 64-picosecond optical pulses and a 2/sup n/ prime code of n=3. A distinguishing feature of this experiment is that the modulation of ultrashort optical clock stream by electrical data is realized without using any optical intensity modulator at each transmitter. Moreover, only low-cost standard optical 2/spl times/2 couplers and fiber delay lines are employed to implement all-serial encoders and decoders for a 2/sup n/ prime code. As a result, this new system is more cost- and power-effective than a conventional AO-CDM system.

We propose a new family of 2/sup n/ codes, called 2/sup n/ extended prime codes, for all-optical code-division multiple access (CDMA) networks. These 2/sup n/ codes are derived from "extended" prime codes such that their cross-correlation functions are not greater than one, as opposed to two in previously proposed 2/sup n/ prime codes. As a result, a larger number of active users can now be supported by the new codes in all-optical CDMA networks where optical multiple-access interference is a major performance limiting factor. Moreover, power-efficient, waveguide-integrable all-serial coding and correlating configurations proposed for the 2/sup n/ prime codes can still be employed by the new codes.

To reduce four-wave-mixing crosstalk in high-capacity, long-haul, repeaterless, wavelength-division multiplexing (WDM) lightwave systems, the use of unequally spaced channels has been proposed. Instead of being solved by integer linear programming, the unequal-spaced channel-allocation problem is treated by constructing suitable optical orthogonal codes in optical code-division multiple-access (CDMA). An "algebraic" framework and three algorithms on finding the frequency locations of unequally spaced WDM channels are introduced, where the constructions are based on generating optical CDMA codewords with a predetermined pulse separation and "aperiodic" autocorrelation sidelobes no greater than one. The algorithms potentially provide a fast and simple alternative to solve the problem, besides the proposed computer-search method.

A "multiwavelength" scheme has been proposed to support large numbers of subscribers and simultaneous users in optical code-division multiple-access (CDMA) networks without using very large bandwidth expansion or the need of complicated and not-yet-feasible optical processing. In this paper, multiwavelength optical orthogonal codes (MWOOCs), which consist of two-dimensional codewords (or matrices) with every pulse of a codeword encoded in a distinct wavelength, are constructed for this scheme. MWOOCs have larger cardinality than the one-dimensional codes used in the hybrid wavelength-division multiple-access (WDMA) and CDMA scheme. With the same hardware configuration, our analysis shows that the multiwavelength scheme, in general, performs better than the hybrid scheme, particularly when the traffic load is heavy. However, if a central controller (i.e., under the best scenario) is used to uniformly distribute all available wavelengths to simultaneous users in the hybrid scheme, both schemes have comparable performance for a medium traffic load and the hybrid scheme can theoretically achieve error-free transmission when the load is light. In addition, using multiple wavelengths, the requirements of fiber ribbons and multiple stars in temporal/spatial optical CDMA networks are eliminated.

Recent study shows that optical code-division multiple-access (CDMA) networks cannot be evaluated or designed by only considering the performance (i.e., correlation properties) of the optical pseudo-orthogonal codes selected. The structures of optical encoders and decoders are another important factors to consider and are needed to coordinate with the selected optical codes as much as possible. A special family of 2/sup n/ codes, so-called 2/sup n/ prime-sequence codes, is constructed. A general theorem on the cardinality of the new codes is provided. The properties and performance of the codes are also studied. Since these codes pose the algebraic properties of both prime-sequence and 2/sup n/ codes, new optical encoding and decoding structures are designed to optimize the system parameters (e.g., power budget and cost) of these optical CDMA networks. This new configuration is particularly attractive for ultrafast optical processing and waveguide implementation for tile future high-capacity, low-loss, all-optical CDMA networks.

An optical orthogonal signature pattern code (OOSPC) is a collection of (0,1) two-dimensional (2-D) patterns with good correlation properties (i.e., high autocorrelation peaks with low sidelobes, and low cross-correlation functions). Such codes find applications, for example, to parallelly transmit and access images in "multicore-fiber" code-division multiple-access (CDMA) networks. Up to now all work on OOSPCs has been based on an assumption that at most one pulse per column or one pulse per row and column is allowed in each two-dimensional pattern. However, this restriction may not be required in such multiple-access networks if timing information can be extracted from other means, rather than from the autocorrelation function. A new class of OOSPCs is constructed without the restriction. The relationships between two-dimensional binary discrete auto- and cross-correlation arrays and their corresponding "sets" for OOSPCs are first developed. In addition, new bounds on the size of this special class of OOSPCs are derived. Afterwards, four algebraic techniques for constructing these new codes are investigated. Among these constructions, some of them achieve the upper bounds with equality and are thus optimal. Finally, the codes generated from some constructions satisfy the restriction of at most one pulse per row or column and hence can be used in applications requiring, for example, frequency-hopping patterns.

An optical orthogonal code (OOC), intended for optical code-division multiple-access (CDMA), is a collection of (0,1) sequences with good correlation properties (i.e., high autocorrelation peaks with low sidelobes, and low crosscorrelation values). A "serial" coding architecture for optical CDMA has been introduced and found particularly suitable for waveguide implementation with high power efficiency. Since the optical encoder is made of a serial combination of 2/spl times/2 optical switches with n of them being biased to a "mix-split" state, the generated OOC's, so-called 2/sup n/ codes, pose with a symmetric delay-distribution property. Various techniques for algebraically constructing such 2/sup n/ codes of weight 4 are presented. The upper bounds on the cardinalities of these codes are also provided.>

A phase-noise-cancelled coherent subcarrier fiberoptic communication system that uses integrated-optic Mach-Zehnder waveguide modulators for frequency-pair encoding is introduced. Optical double-sideband suppressed-carrier signals with the same random phase noise and information encoded in the frequency separation of the two sidebands are generated by properly biasing the modulator. Phase noise is eliminated at the receiver by a nonlinear operation on the two sidebands. System performance analysis with numerical examples, taking into account the modulation index and laser normalized linewidth, is presented.

Synchronous code-division multiple access (S/CDMA) is investigated for fiber-optic local area networks. It is shown that the large bandwidth expansion required by spread-spectrum techniques, such as CDMA, can be accommodated by using a fiber-optic channel for transmission and incoherent optical signal processing for code generation and correlation. Prime sequence codes, previously developed for a fiber-optic network using (asynchronous) CDMA, are modified to fit a synchronous transmission format. A performance comparison of CDMA and S/CDMA systems reveals that S/CDMA can accommodate a larger number of subscribers and more simultaneous users than CDMA. An environment for which S/CDMA would be suited is discussed.>