Konstantinos Paparrizos

Enhancing performance of the IEEE 802.11 distributed coordination function via packet bursting

During the past few years, wireless local area networks (WLANs) have become extremely popular. The IEEE 802.11 protocol is the dominating standard for WLANs employing the distributed coordination function (DCF) as its essential medium access control (MAC) mechanism. This paper presents a simple and accurate analysis using Markov chain modelling to compute IEEE 802.11 DCF performance, in the absence of hidden stations and transmission errors. This mathematical analysis calculates in addition to the throughput efficiency, the average packet delay and the packet drop probability for both the basic access and RTS/CTS medium access schemes. The derived analysis, which takes into account packet retry limits, is validated by comparison with OPNET simulation results. The mathematical model is used to study the effectiveness of the RTS/CTS scheme at high data rates and the performance improvements of transmitting a burst of packets after winning the contention for medium access. Packet bursting considerably increases both throughput and packet delay performance but lowers the short-term fairness on medium access.

Packet Delay Metrics for IEEE 802.11 Distributed Coordination Function

In this paper, we introduce a comprehensive packet delay analysis for wireless networks based on IEEE 802.11 Distributed Coordination Function (DCF). We develop mathematical models that calculate a set of packet delay metrics, namely a) the average packet delay for successfully transmitted packets, b) the average packet delay of successfully transmitted packets experiencing a specific number of collisions, c) the average packet drop time, d) the delay jitter and e) the delay distribution by computing the probability of a packet to be successfully transmitted experiencing delay time lower than a given value. All the developed models are based on calculating station’s delay time at the transmission slot(s) plus the average time that station defers at backoff slots before successful transmission. The mathematical models are simple, computationally fast and can be used to build admission control algorithms. Simulation results show that our proposed mathematical analysis is highly accurate.

Packet delay distribution of the IEEE 802.11 distributed coordination function

The paper studies the packet delay distribution of the IEEE 802.11 distributed coordination function (DCF) protocol. DCF treats packets in an unfair manner. Results indicate that, in large networks, most packets have very low time delays, some packets have delays close to the average value and a small number of packets experience extremely high delays. We study the DCF delay distribution by developing a mathematical model that calculates the important properties of the constituent curves of the delay distribution curve, namely the probability that a packet is successfully transmitted from a particular backoff stage and the average delay of the successfully transmitted packets from this backoff stage. The model is simple, gives an insight into the internal mechanisms of DCF and applies to both basic and RTS/CTS access mechanisms. The accuracy of the analytical model is verified by simulations. Analytical results are presented that explore the effect of network size and of the initial contention window size on the fairness of DCF regarding the distribution of packet delays.

A new efficient primal dual simplex algorithm

The purpose of this paper is to present a revised primal dual simplex algorithm (RPDSA) for linear programming problems. RPDSA has interesting theoretical properties. The advantages of the new algorithm are the simplicity of implementation, low computational overhead and surprisingly good computational performance. The algorithm can be combined with interior point methods to move from an interior point to a basic optimal solution. The new algorithm always proved to be more efficient than the classical simplex algorithm on our test problems. Numerical experiments on randomly generated sparse linear problems are presented to verify the practical value of RPDSA. The results are very promising. In particular, they reveal that RPDSA is up to 146 times faster in terms of number of iterations and 94 times faster in terms of CPU time than the original simplex algorithm (SA) on randomly generated problems of size 1200 × 1200 and density 2.5%.

An efficient simplex type algorithm for sparse and dense linear programs

We present a big-M method for solving general linear programming problems with a recently developed exterior point simplex algorithm (EPSA). We also provide intuitive motivations to EPSA. We describe major steps in implementing EPSA for large-scale linear programming. Our implementation was carried out under the MATLAB environment. This algorithm seems to be more efficient than the classical primal simplex algorithm (PSA), employing Dantzigs rule. Preliminary computational studies on randomly generated sparse and dense small and medium linear programs support this belief. In particular, EPSA is up to 10 times faster than PSA on medium size linear programs. This translates into corresponding savings in CPU time. Although the computational effort required in each step of EPSA requires more time compared to an iteration step of PSA, the improvement of EPSA comes from the fact that it requires adequately less iterations than PSA. Moreover, as the problem size increases and the problem density decreases, EPSA gets relatively faster. 2002 Elsevier Science B.V. All rights reserved.

E-Learning Technologies: Employing Matlab Web Server to Facilitate the Education of Mathematical Programming.

This paper presents new web-based educational software (webNetPro) for Linear Network Programming. It includes many algorithms for Network Optimization problems, such as shortest path problems, minimum spanning tree problems, maximum flow problems and other search algorithms. Therefore, webNetPro can assist the teaching process of courses such as Graph Theory or Network Optimization, Algorithms or Data Structures. Since the Matlab web server is part of the underlying infrastructure, the end user need not use Matlab directly. The approach exploits transparently the benefits of the Matlab programming environment through a straightforward web interface. Furthermore, webNetPro is platform independent. Thus, it can be viewed as a powerful supplement to traditional instruction techniques and can be used without significant difficulties in distance education. Benefits and drawbacks are thoroughly described in order to support the significance of this tool in distance learning.

An exterior point simplex algorithm for (general) linear programming problems

We present an exterior point simplex type algorithm that possesses a new monotonic property. A dual feasible basic solution is required to start with. Intermediate solutions are neither primal nor dual feasible. Cycling-free pivoting rules and an exponentional example are presented.

An exterior simplex type algorithm for the Minimum Cost Network Flow Problem

In this paper a new Network Exterior Point Simplex Algorithm (NEPSA) for the Minimum Cost Network Flow Problem (MCNFP) is analytically presented. NEPSA belongs to a special simplex type category and is a modification of the classical network simplex algorithm. The main idea of the algorithm is to compute two flows. One flow is basic but not always feasible and the other is feasible but not always basic. A complete proof of correctness for the proposed algorithm is also presented. Moreover, the computational behavior of NEPSA is shown by an empirical study carried out for randomly generated sparse instances created by the well-known GRIDGEN network problem generator.

Computational experience with exterior point algorithms for the transportation problem

An experimental computational study to compare the classical primal simplex algorithm and the exterior point algorithms for the transportation problem (TP) is presented. Totally, four algorithms are compared on uniformly randomly generated test problems. The results are very encouraging for one of the competitive algorithms. In particular, a dual forest exterior point algorithm is on average up to 4.5 times faster than network simplex algorithm on TPs of size 300x300 and for all classes. This result leads into corresponding savings in computational time. From the computational performance we conclude that as the problem size increases, exterior point algorithm get relatively faster.

Delay Distribution Analysis of the RTS/CTS mechanism of IEEE 802.11

A simple and accurate packet delay distribution analysis is presented for the RTS/CTS access scheme of the IEEE 802.11 distributed coordination function (DCF). The packet delay distribution is effectively studied by developing an analytical model that calculates the probability that a packet is successfully transmitted after experiencing a delay time equal to a given value. The model uses simple mathematical relations and provides accurate delay distribution curves. The accuracy of the analytical model is verified by simulations