C.K. Un

Performance analysis of Markovian polling systems with single buffers

Abstract In this paper, we analyze the Markovian polling system with single buffers, asymmetric arrival rates, service times, and switchover times. A virtual buffer model is introduced to derive the relationship of the joint generating function for the queue length of each station at a polling instant. The Laplace-Stieltjes transforms of the cycle time and the intervisit time are obtained from the marginal generating function. We analyze the cyclic, load-oriented-priority, and symmetric random polling schemes which are classified by adjusting the transition probabilities, and compare the merits and demerits of each scheme for the performance measures. In particular, we prove that the mean queue lengths at the polling instants are the same for all stations in case of the load-oriented-priority polling scheme for the buffer relaxation system in which a new message is stored as soon as the transmission of the message currently in the buffer is initiated.

A new distributed location management algorithm for broadband personal communication networks

We propose an efficient location management algorithm for broadband personal communication networks (PCNs) which reduces the complexity of network signaling and delay due to database access when locating a portable station (PS). Previous works on updating and search algorithms for personal communications have concentrated on the access delay due to signaling network traffic only. But, with the advances in fiber optic technology, it will be possible to provide the capacity of signaling network on the order of several Mbps in broadband PCNs. Thus it may be more important to simplify the signaling procedure and reduce the number of database accesses for mobility management at the cost of some increase of signaling traffic. In the proposed algorithm, we first remove all visitor location registers (VLRs) in order to achieve the simplicity of signaling procedure. Thus, only home databases for mobility management are distributed in a global network, and each database keeps only the current locations of PSs initially registered in its management area. It can solve the problem of requiring an impractically large database in a centralized database structure by distributing home location registers (HLRs) while maintaining the simplicity of mobility management and reducing database loads. This algorithm has a simple signaling procedure, and reduces database operations when compared to hierarchical algorithms as network switching hierarchy increases or relocation and calling rate increase. >

Performance comparison of nonpreemptive and preemptive priority queueing strategies in ATM packet switch with input buffers

Abstract In this paper, we study the performances of nonpreemptive and preemptive priority queueing strategies in an asynchronous transfer mode (ATM) packet switch with input queues. We first investigate the priority effects on the ATM packet switch, and consider some variations of the priority queueing strategy. Next, we analyze the performances of delay and throughput for the priority queueing disciplines and obtain closed-form solutions. We then compare the performances of the nonpreemptive and preemptive schemes. From the results, we find that, in the switching system with the preemptive priority queueing strategy, the head-of-line (HOL) service time of a low-priority packet may be overlapped with the HOL service time of a high-priority packet when the HOL preemption occurs. Therefore, unlike a general queueing system, the preemptive priority queueing strategy can improve the performance of low- and high-priority packets. In addition, with the closed-form solutions obtained, we can reduce the computational complexity or eliminate heuristic adjustments to analyze the performance of the switching system with the priority queueing strategy.

Capacity of Wireless Dynamic TDMA and Media Access Control for Packetized Voice/data Integration

We obtain the capacity of wireless dynamic TDMA for packetized speech and investigate its stability characteristics depending upon the ratio between request slots and message slots. We analyze this system based on an imbedded Markov chain approximation and validate its performance by simulation. We also propose a new media access control for voice/data integration and investigate its performance by simulation

Performance of contention processing schemes for WDM networks with control channels

Abstract In this paper, four contention processing schemes are studied for a single-hop wavelength division multiplexing (WDM) network with control channels. In order to improve the performance of the network while avoiding the electronic bottleneck, we propose two novel contention processing schemes: local collision blocking (LCB) and local collision blocking-contention resolution (LCB-CR) schemes. We also consider two other schemes: pure contention and contention resolution unit-only (CRU-only) as the lower and the upper performance bounds, respectively. Applying these four schemes to the delayed-ALOHA access protocol (Mehravari, 1990), we analyze the performances of these four schemes under a finite population model and asymmetric conditions. We also take account of receiver collision in the analysis. From the comparison, we find that the proposed schemes are effective in achieving significant performance improvement with moderate system complexity. In order to design the system properly, we consider the effects of various parameters by numerical results.

Performance analysis of an input queueing ATM switch with two priority classes

Abstract In this paper, we study an input queueing asynchronous transfer mode (ATM) switch with the preemptive priority scheme. Up to now, many researchers have studied this switch, but their models do not yield accurate results. Therefore, we propose another model and analyze the performance of the switch for two different arrival processes, that is, the independent priority arrival process (IPAP) and the dependent priority arrival process (DPAP). In this paper, we model a virtual head of line (HOL) queue like the previous works, and analyze it using a three-dimensional Markov chain. As for each input queue, we model it as a two-class Geom x /Geom/1 queue with preemptive priority. By using this model, we obtain the mean queueing delay of the low priority class and the maximum throughput of the switch. In case of the IPAP, we compare our result with simulation and those of the previous works to show that our model yields more accurate results. In addition, we compare the performances of those two arrival processes to illustrate their difference.

Performance analysis of input queue with combination of push-out and state-dependent scheduling schemes for nonblocking ATM switches

Abstract In this paper, assuming two classes of input traffic, we analyze a combination of push-out and general state-dependent scheduling schemes applied to input queues of nonblocking ATM switches with the speed-up factor ≥ 1. The analysis is done in two steps in the discrete-time domain, using an embedded semi-Markov process (SMP) concept (Takagi, 1993). In numerical examples for some specific scheduling schemes, we compare the performances and show that reasonable performance differences both in loss probabilities and in waiting times can be achieved between two classes over a wide range of the offered load. We also show that all the priority schemes considered in this paper allow higher utilization of the switch than the non priority one, and among them the queue length threshold (QLT) scheduling scheme gets the highest utilization. Lastly, we note that although the analysis assumes an infinitely large size switch, the result gives a quite accurate approximation for finite size switches with around one hundred input/output ports.

Performance analysis of two fanout splitting schemes for a multicast packet switch with capacity m

Abstract A multicast packet switch provides an efficient transport method which can connect any input port to any subset of output ports. Each input packet which is copied and destined to a subset of output ports usually experiences contention with other copy packets in a space-division packet switch, thus requiring some selection rule (queueing discipline) among the contending copies. This paper studies the performance of a multicast packet switch with capacity m which adopts one of two fanout splitting disciplines, i.e., the first-come-first-served (FCFS) head-of-line (HOL) service discipline and the random HOL service discipline. First, we analyze the service time distribution of a copy packet and an input packet. Using this distribution, the saturation throughput of the switch, mean delay and input queue length distribution are obtained. By numerical analysis and simulation, we show that the FCFS HOL service outperforms the random HOL service.

Performance analysis of cut-through buffered banyan networks with finite buffer size

Abstract In this paper, we present an analytic model to estimate the performance of cut-through buffered banyan networks with finite buffer size. Two conflict resolution policies are considered in order to resolve conflicts when two packets are destined to the same output link, and their performances are compared. Our analytic model enables analysis of the networks in which buffers are unevenly distributed, i.e., each stage has a different buffer size. It is shown that if buffers are properly distributed in the network, higher throughput and lower delay are possible, although the improvement is considered small. Finally, in order to validate our model, some analytic results are compared to simulation.

Research: Performance of dilated banyan network with back-pressure mechanism

A dilated banyan network is one solution that can overcome the internal blocking problem in a banyan network. However, tremendous routing capacity is wasted in the dilated banyan network because d - 1 links of each input port in the first stage are unused. In this paper, we propose a dilated banyan network with back-pressure mechanism to fully utilize the wasted routing capacity. The proposed switch has input and output queues, and employs a priority scheme to prevent out-of-sequence. We also design a non-blocking element in order to reduce the hardware complexity of the dilated banyan network. The performance of the proposed switch is analyzed under uniform traffic assumption. Numerical and simulation results show that the proposed switch yields a significant improvement of the maximum throughput as compared to that of the pure dilated banyan network. Moreover, with dilation degree of d = 3, the maximum throughput of about 1 is achieved.