The main concern in designing the switching fabrics used in asynchronous transfer mode (ATM) technology are speed, throughput, delay and variance of delay for a given bandwidth. We analyze the performance by simulations of ATM switch based on Banyan network in the uniform traffic condition. We compare the analytical results obtained from the three-state model by Yan and Jenq (see MS Thesis, Dept. of Electrical and Computer Engineering, Portland State University, Portland, Oregon, 1995) to the simulation results. We propose non-blocking first stage (NBFS) to increase the throughput. In the NBFS scheme, internal blocking in the switching element (SE) of the first stage is avoided. We extend the idea of the priority scheme proposed by Yan and Jenq by proposing an enhanced priority (EP) scheme. In the EP scheme, we give priority to the packet blocked the greater number of times. We simulated the switch with a combined NBFS and EP scheme. The results shows that the throughput increases, and the variance of delay decreases significantly. We simulated the single buffer Banyan network with mixed voice and data traffic We simulated the Banyan network with the proposed double buffer switching element to increase the throughput and decrease the delay and variance of delay, of delay sensitive voice packet in a mixed traffic condition.

We propose a new model for the performance analysis of an ATM switching fabric based on a single-buffered Banyan network. In this model, we use a three-state, i.e., "empty", "new" and "blocked" Markov chain model to describe the behavior of the buffer within a switching element. In addition to traditional statistical analysis including throughput and delay, we also examine the delay variation. Performance results show that the proposed model is more accurate in describing the switch behavior under a uniform traffic environment in comparison with the "two-state" Markov chain model developed by Jenq (1983) and Yoon (1990). Based on the "three-state" model, we study a packet priority scheme which gives the blocked packet higher priority to be routed forward during contention. For a fully loaded six-stage switch, the analytical model predicts that the delay variation can be reduced by 35%, while the simulation shows a 22% reduction.

The overwhelming majority of the digital signal processing theories developed so far have been concentrated on uniformly sampled signals. In this paper we develop a systematic approach to compute the digital spectra of a class of digital signals obtained by nonuniformly sampling analog signals. This development is motivated by the need to characterize the spectral property of a digitally synthesized sinusoidal waveform. The full representation of the spectral components of a digitally synthesized sinusoidal signal is derived. From that representation, we prove some important properties which provide great insight into the spectral characteristics of a digitally synthesized sinusoid.

Banyan networks are being proposed for interconnecting memory and processor modules in multiprocessor systems as well as for packet switching in communication networks. This paper describes an analysis of the performance of a packet switch based on a single-buffered Banyan network. A model of a single-buffered Banyan network provides results on the throughput, delay, and internal blocking. Results of this model are combined with models of the buffer controller (finite and infinite buffers). It is shown that for balanced loads, the switching delay is low for loads below maximum throughput (about 45 percent per input link) and the blocking at the input buffer controller is low for reasonable buffer sizes.

In this paper, we study the adder overflow problem of a second order IIR filter in a more direct way by considering the probability of state overflow. Equations which relate the scale factor and the state overflow probability are derived. This scaling policy is applicable to different transfer function properties and various implementation structures, e.g., the canonical form realization, the uniform grid form realization, and the low sensitivity form realization. Numerical examples and simulation results covering a broad scope of filter implementation problems are presented. We also give quantitative comparisons for various schemes of handling the overflow problem. The results of this paper will offer the filter designer a realistic and efficient way to deal with the forced response overflow problem.

The calculation of convolution (noncyclic) of two finite sequences is a fundamental operation in digital signal processing. In this paper we derive two digital convolution algorithms which are efficient in terms of number of multiplications and also structurally simple, and hence suitable for implementations in a tree machine or in a VLSI chip where a simple geometry of circuit layout is highly desirable.

Multiple access methods constitute an important subject in the design of distributed computer communication systems. A technique which has attracted considerable attention is the slotted ALOHA random access scheme. In this paper, based on some analytical properties of the slotted ALOHA system [9], we establish the optimality property of the multiple control limit policy among the class of multiple-action policies. The previously reported control limit policy [6] is just a special case. We then derive the optimal retransmission control (ORC) policy which is optimal among the whole class of stationary policies. The superiority of the ORC policy is also illustrated by numerical examples. The performance of the ORC policy for various potential input traffic intensity is presented.

An approximation algorithm is developed for calculating transient mean and variance of the queue size of a discrete queueing system with independent general arrivals and geometric output processes. The algorithm makes no assumption on the form of the distribution function of the arrival process and only the first three moments of the arrival process are required. Three numerical examples are presented.

In this correspondence, we investigate an intuitive conjecture on data communication studies that a larger intersymbol interference (ISI) would result in a higher error probability. It is proved that if: 1) the eye is open, and 2) the additive noise has a symmetric probability density function and its complementary distribution function is concave upward (i.e., convex\cup) on the positive real line, then the system with a larger ISI will have a higher probability of error.

In this correspondence we prove the following properties of the stability of the finite population model of slotted ALOHA systems. 1) The input-output packet flow balance principle [1] and the concept of expected drift [2] used for the stability analysis of the slotted ALOHA system are mathemtaically equivalent. 2) The slotted ALOHA system can only have either one stable equilibrium point or three equilibrium points, with the first one and the third one stable and the second one unstable. 3) If the retransmission probability (which is assumed to be greater than the probability of new packet generation) is less than or equal to2/N, whereNis the total number of active users in the system, then the system has only one stable equilibrium point.

This paper considers the effect of the carrier phase error on the probability of error in a binary quadrature amplitude modulation (QAM) data transmission system in the presence of intersymbol interference and additive Gaussian noise. The analysis shows that there exists a threshold value of the carrier phase error such that the error probability remains almost unchanged as long as the carrier phase error is less than that threshold value, but increases rapidly once the threshold is exceeded. Good agreement between theory and computer simulations is obtained. Some practical implications of this threshold effect are discussed.

This paper considers the design of a tapped delay line (TDL) equalizer from an error probability point of view. We define a new class of TDL equalizers, which includes the zero-forcing equalizer (ZFE) as a special case. Some optimal properties of the new equalizers are presented and their relationships to mean-square equalizers (MSE's) and ZFE are discussed. It is shown that, for a certain class of additive noise, some of the new equalizers result in a probability of error lower than that of the MSE.

The effect of intersymbol interference and additive noise on the performance of a digital data transmission system is considered. The data sequence is assumed to be independent and equiprobable. The additive noise is independent of the signal but is not restricted to be Gaussian. A simple lower bound, an upper bound, and a simple approximation to the upper bound, on the probability of error are derived. The approximation to the upper bound is twice the lower bound; hence, either can be taken as an approximation to the actual error probability.