We consider frame synchronization for fixed frame length communication system with periodically embedded sync words. In the past few decades, numerous researchers have made significant contributions in various aspects. However, it is almost surprising that a scrutinization over the literature reveals to us an important common assumption: frame synchronization decision-making through single-frame observation. Intuition tells us that an enlarged multiple-frame observation set bears the potential to reduce sync word identification ambiguities and also to combat multi-path fading through ¿frame diversity¿. The objective of this work is thus to characterize the performance improvement of using multiple-frame decision rules rather than the single-frame decision rules considered in the previous work. We select the work of Lui and Tan as our standing point, and have conducted a comparative study in a very comprehensive manner. In particular, we shall show that the benefits of multipleframe based decision rules can be substantial. For example, in many cases, the performance of the correlation rule based on double-frame observation is superior to that of the optimum maximum likelihood rule with single-frame observation. While further enlarging the observation period beyond double frames was found to yield diminishing returns.

Previous work on the bit error rate (BER) performance of differential phase-shift keying (DPSK) including the effects of additive white Gaussian noise (AWGN) and phase noise has concentrated on the delay demodulator with narrow-band intermediate frequency (IF) bandpass filtering (BPF) and sampling detection. No similar analysis has yet been performed for the delay demodulator with wideband IF bandpass filtering and integration detection. Phase noise is an important consideration in coherent optical communication systems and the most widely accepted model is a Brownian motion process. A closed-form BER expression along with detailed Monte Carlo simulation results are presented for the DPSK delay demodulator with wideband IF bandpass filtering and integration detection filtering including phase noise effects using the Brownian motion model. Analytic expressions are also obtained for the moments of the phase noise component of the decision variable. Using these moments, estimates of the phase noise BER floor are produced. It is found that this receiver has noise performance comparable to receivers with narrowband IF bandpass filtering and sampling detectors for signal-to-noise ratio (SNR) and phase noise in the range of practical interest, but with potentially less degradation due to intersymbol interference (ISI).

This paper considers a generic CSMA and CSMA/CD protocol that allows one to unify the nonpersistent and one-persistent protocols as special cases, and solve the characterization problem under the infinite population model. Explicit expressions of the interdeparture time moment generating function, as well as expressions for the channel throughput and the variance of output process, are obtained for both slotted and nonslotted cases. In addition to direct simulation results, we also provide a novel statistical analysis/simulation procedure for the purpose of validating the infinite population model and the analytical results. Both approaches indicate that the infinite population model is quite appropriate for analyzing this generic protocol.

The proposed implementation of two QOS bearer services in ATM networks has generated considerable work on the analysis of implicit policy assignment space priority control buffer access schemes. Recently, there have been more interest directed towards explicit policy assignment space priority schemes because of their increased efficiency and flexibility over implicit assignment schemes. The paper considers an explicit policy assignment space priority control scheme employing the push-out buffer access mechanism. Cell loss probabilities are derived for a traffic model consisting of two service classes. Each service class consists of bursty traffic generated by a multiple number of three-state discrete-time Markov sources. The effect of cell loss probabilities on traffic load and burstiness characteristics, as well as buffer size, is examined. The authors have previously analyzed the performance of an explicit policy assignment partial buffer sharing scheme for the same traffic model. Performance comparisons between the push-out and partial buffer sharing schemes are also made in the present paper.>

The probability distribution of the diameter of a network model consisting of a complete digraph with positive integer random edge costs is considered. Edge costs are chosen independently for each node pair according to common probability distributions for each edge direction. Bounds and some sharp limit results for the diameter distribution in the limit as the number of nodes tends to infinity are derived. In addition, limit bounds are determined for the distribution of the number of hops in a shortest path between two arbitrary nodes in the graph. Numerical examples are presented to illustrate these results.>

A theoretical analysis of the cell loss probability performance of an explicit priority assignment partial buffer sharing scheme for asynchronous transfer mode (ATM) networks is provided. A two-service-class traffic model is used in which each service class consists of bursty traffic generated by a multiple number of discrete-time Markov sources. The effect of cell loss probability on both traffic load and burstiness characteristics, as well as on system parameters such as buffer size and buffer sharing threshold, is examined. These results are useful for system design. The many examples examined yield no significant priority separation for either high- or low-priority cells. This is particularly desirable for fulfilling the objective of two quality of services (QOS) bearer services to provide only two cell loss probability levels of service.>

The probability distribution of the shortest path cost from a source node to an arbitrary destination node is considered for a random network model consisting of a complete digraph with positive integer random edge costs. Edge costs are chosen according to a common probability distribution for each direction. For this model, the joint distribution of the number of nodes which have a given sequence of shortest path costs from an arbitrary source node is determined explicitly. An expression is then obtained for the distribution of the shortest path cost between two arbitrary nodes using this joint distribution. The main result is the derivation of tight bounds and a sharp limit result for the distribution of the shortest path cost as the number of nodes tends to infinity. Numerical examples are presented to illustrate these results.>

The average throughput, average delay, and the first exit time performances are determined for a slotted two-packet replication Aloha protocol in a fading communications channel environment using a two-state Gilbert (1960) channel error model with memory. Comparison to slotted Aloha without replication shows that a smaller total average delay can be achieved by using two-packet replication. Average throughput is improved in some cases. However replication reduces the stability of the system. The Markov model for the slotted replication Aloha system is discussed, and the stability is quantified in terms of the expected first exit time from the safe channel operating region.>

The concepts of expected drift and first exit time are applied to address the stability issue of a generic p-persistent carrier sense multiple access with collision detection (CSMA/CD) protocol that includes the nonpersistent and 1-persistent protocols as special cases. Both slotted and nonslotted operations are treated. By utilizing the zero-drift condition as an input-output balance condition, the formidable tasks of specifying the channel load lines and equilibrium contours are avoided in this approach when assessing the stability performance of a persistent CSMA/CD system. Also characterized is the packet output of the process of the generic p-persistent CSMA/CD protocol under a finite population model. Explicit expressions are provided for calculating the first moment of the packet interdeparture time so that the channel throughput and packet delay can be easily evaluated.>

The performance of the slotted ALOHA protocol in a fading multipath communications environment is determined using a two-state Gilbert-channel error model with memory. Average throughput and delay evaluations are performed for a finite-user population model. The stability of the protocol is determined by evaluating the average first exit time (FET) from the safe operating region. Gilbert clustered-channel error models having clustered error burst behavior are considered as representative of fading multipath channels. The average throughput, delay, and FET performance for clustered channels are shown to be significantly worse than that for memoryless channels with the same packet error rate.>

Two techniques that prioritize the common channel acknowledgment traffic are proposed and analyzed for packet switched cable systems that use carrier-sense multiple-access (CSMA) protocols. Closed-form expressions for the packet interdeparture time moment generating functions of both priority acknowledgement schemes are obtained, under a Poisson arrival model, to facilitate the performance comparisons with respect to the channel throughput and output variation coefficient. A generic p-persistent carrier-sense protocol is utilized to unify the treatments of the nonpersistent and 1-persistent CSMA protocols. A statistical test procedure is also introduced to evaluate the appropriateness of the Poisson arrival model. Slotted operation and nonslotted operation are treated, along with the use of collision detection (CD).>

A novel connection request procedure for a land mobile satellite system (LMSS) is described. The procedure is based on a binary-feedback free-access collision resolution algorithm (CRA) operating over a number of parallel channels. This protocol is shown to achieve a much higher stable throughput than the previously recommended modified slotted ALOHA protocol for making connection requests. The complexity involved in implementing such a protocol is inherently low. Furthermore, this protocol is superior to other multiple-channel obvious blocked-access CRA protocols. By interleaving the time slots among individual channels, it can be applied to multiple-channel land mobile satellite type systems. Simulation results are obtained that demonstrate stable performance up to capacity. Moreover, the delay throughput characteristic is extremely flat over a wide range of traffic input rates.>

An approximate maximum-likelihood rule (ML) is derived for joint symbol and frame synchronization with pulse-position modulation, Simulation results show that this approximate rule performs virtually as well as the optimum ML rule over a wide range of practical signal-to-noise ratios. Yet it is substantially simpler to implement.

The problem of locating a periodically inserted frame synchronization pattern in random data for aM-ary digital communication system operating over the additive white Gaussian noise channel is considered. The optimum maximum-likelihood decision rule, high signal-to-noise approximate maximum likelihood decision rule, and ordinary correlation decision rule for frame synchronization are derived for both coherent and noncoherent phase demodulation. A general lower bound on synchronization probability is derived for the coherent correlation rule. Monte Carlo computer simulations of all three decision rules, along with evaluations of the lower bound for the coherent correlation rule, were performed for the coherent MPSK, coherent, and noncoherentMary orthogonal, and 16 QAM signaling schemes. These results show that in each case the high signal-to-noise maximum-likelihood rules have a performance nearly equal to that of the maximum-likelihood rules over a wide range of practically interesting signal-to-noise ratios (SNR's). These high SNR decision rules also provide significant performance improvement over the simple correlation rules. Moreover, they are much simpler to implement than the maximum-likelihood decision rules and, in fact, are no more complex than the correlation rules.

The problem of locating a periodically inserted frame synchronization pattern in random data is considered for a directdetection optical communication channel. Both pulse-position modulation and on-off keying with known symbol synchronization are considered. The high signal-to-noise ratio approximation of the maximum likelihood decision rule is derived for each of these cases. Analytical bounds and simulation results on synchronization probability show that the high SNR ML rule has a performance approaching that of the optimum ML rule over a wide range of signal-to-noise ratios. This decision rule also provides a substantial performance improvement over the correlator rule with virtually no additional implementation complexity.

The explicit evaluation of the absolute error rate-distortion function has been obtained previously for a class of i.i.d, sources with probability densities supported on the entire real line. The technique is extended to various i.i.d, sources with densities supported on a proper subset of the real line, such as uniform and one-sided exponential densities. Comparisons of these rate-distortion functions with their Shannon lower bounds are also given.

A tree coding theorem is established for discrete-time abstract alphabet stationary block-ergodic sources under a single-letter fidelity criterion. The proof of this theorem uses recent results from the theory of branching processes with random environments. The relation to previous work in tree coding is also discussed.

We define a symmetric modular source to be a source-user pair consisting of an equiprobable discrete memoryless source together with a modular context-dependent fidelity criterion. For this source-user pair we establish the existence of block codes of fixed rateRgreater thanR(D)with average distortion that decreases towardsDat a doubly exponential rate in block length.

The explicit evaluation of the rate-distortion functionR(D)for continnous-amplitude sources when it is strictly greater than its Shannon lower bound has been previously considered a formidable task. This is the case for independent identically distributed (i.i.d.) Gaussian sources under an absolute-magnitude distortion criterion. In this paper, we provide a method that yields the explicit evaluation ofR(D)for this case. This method is then shown to be applicable to the evaluation of the absolute-magnitude-criterionR(D)for a large class of i.i.d, sources having probability densities with constrained tail decay. Other applications are discussed.

We consider a class of strictly stationary binary sources\{u_n, n \geq 1\}, characterized by joint conditional probability mass functions \begin{equation} Q_n(U_n = k_n \mid \theta) = \prod_{i=1}^n \theta^{1-k_i} (1 - \theta)^{k_i} \end{equation} for alln \geq 1, where\thetais an r.v. on