Emmanuel N. Protonotarios

Performance Analysis of the Selective Repeat ARQ Protocol

Delay, queue length, and throughput are the main performance characteristics of a data transmission system with ARQ (automaticrepeat-request) error control. Various protocols which have been proposed use, as a part or as a whole, the basic selective-repeat ARQ protocol. Their performance analysis has, however, been restricted to the throughput characteristics. An easily applicable method to analyze this basic protocol has not appeared in the literature so far. In this paper, two different methods for the queue length and delay analysis are presented. The system is modeled as a discrete time queue with infinite buffer storage. Transmission errors are considered to be independent, and block arrivals may follow an arbitrary interarrival time distribution. The first method uses an exact Markov state model, on which the theory of absorbing and ergodic Markov chains is applied, and leads to a computational algorithm. The second method, which is based on a specific assumption, uses a substantially simpler stochastic model and results in equations which are easily solved by means of iterative computation. In the case of geometrically distributed interarrival times, simple analytical formulas are extracted. The results are compared to the exact ones (that is, those obtained by the first method or by extended simulation runs) and surprising agreement is observed.

Queueing Analysis of Buffered CSMA/CD Protocols

In this paper, we propose a new queueing model appropriate for the analysis of a buffered CSMA/CD protocol. We assume that each user has a finite buffer capacity. The system, in general, can be modeled as a multidimensional semi-Markov chain. The enormous state space for the complete description is reduced by appropriate exploitation of the features of the system, e.g., fairness towards all users, dependence of the probability of successful transmission only on the total number of busy users, etc. Using a further assumption we construct a two-dimensional semi-Markov chain, examining the state of the system immediately after the successful transmission of packets. The obtained solution is extremely accurate and exhibits excellent agreement with simulation results, which corroborates the correctness of our method.

Simulative analysis of optimal resource allocation and routing in IBCN's

The problem of resource allocation for future integrated broadband communication networks (IBCNs) is addressed. It mainly involves resource allocation at the connection level. The resource allocation problem is decomposed into the following interdependent tasks: given that a network can accommodate the bandwidth demand of a call request, determine a route for the corresponding asynchronous transfer mode (ATM) virtual connection; and allocate bandwidth, i.e. links inside the trunks of the chosen route, to this connection according to predefined limits on bandwidth use by various service calls. Various link allocation schemes combined with routing algorithms are examined. Their performance in terms of service call blocking is evaluated using a software package developed, for that purpose. It is shown that the traditional complete sharing (CS) and complete partitioning (CP) policies are not adequate for IBCNs. Movable boundary (MB) policies are more flexible and present near-optimal performance when access of broadband service to narrowband service resources is allowed and suitable routing algorithms are dynamically applied. >

Quality of service requirements in ATM-based B-ISDNs

Abstract The purpose of this paper is to present the Quality of Service (QOS) requirements of bearer services in an ATM based B-ISDN environment. Asynchronous Transfer Mode (ATM) is a specific packet-oriented switching and multiplexing technique for carrying future broadband telecommunication services. An extremely wide variety of services with different traffic characteristics and performance requirements is expected to be supported by the ATM based B-ISDN. In this paper, the ATM QOS parameters are studied on a per service basis. For this purpose a number of characteristic bearer services has been selected, and the required values for the ATM QOS parameters for each of the selected services have been defined.

Queueing Analysis of Some Buffered Random Multiple Access Schemes

A queueing model appropriate for the analysis of multipleaccess schemes with finite user buffer capacity is proposed. This model can be applied to several slotted random multiple-access schemes, such as the URN, ALOHA, random TDMA channels, etc. The examined system, in general, can be modeled as a multidimensional Markov chain. The enormous state space for its complete description makes numerical analysis intractable. Nevertheless, symmetry properties of the system can be exploited so that the state space will be reduced. In particular, the queue length at one station and the number of busy stations are used to characterize the behavior of the system. In this way, a two-dimensional Markov chain can be constructed and an approximate solution is obtained which is in excellent agreement with simulation results.

An OXC Architecture Suitable for High Density WDM Wavelength Routed Networks

A novel optical crossconnect architecture, capable for tens of terabit throughput and allowing the integration of wavelength and space switching, is presented for both WP and VWP networks. The architecture offers significant advantages in terms of crosstalk performance and node complexity. In the architecture, there are no crossovers and crosspoints. Therefore, the signal-to-crosstalk ratio per wavelength channel can be very high since each channel is processed independently from the others. This will also allow to introduce other all-optical techniques like adding/dropping of OTDM channels as well as optical regeneration. The proposed architecture is wavelength and link modular and particularly suitable for crossconnects with a very large number of wavelength channels. Performance aspects of a single as well as of a cascade of crossconnects have been simulated using a simulation tool.

A transparent MAC method for bandwidth sharing and CDV control at the ATM layer of passive optical networks

In the quest for a way to bring local loop costs down and make B-ISDN affordable to small traffic customers, the passive optical networks (PONs) in combination with ATM (APONs) have been investigated extensively. To achieve the required medium sharing on a time division multiplexing (TDMA) basis, an arbitration mechanism is required in the upstream direction. The resulting multiplexing and concentration, though it is the source of significant savings, it has been plagued by the penalty of high distortion of the traffic profiles at the exit of the PON. The preventive congestion control method of ATM is rather intolerant of such distortions in the multiplexing stages. Arbitration methods which are either semistatic or based on number of arrivals per reservation period, cause very high cell delay variation (CDV). The cell clustering introduced by the arbitration method cannot be distinguished from peak rate violations of the contract by the policing unit. The result is either unwarranted rejections of conforming cells or unacceptably low utilization. The scheme presented herewith maintains a CDV behavior comparable to that of a common centralized FIFO multiplexer. It relies on a meticulous recording of the arrival timing of cells enabling a subsequent allocation of slots which produces almost the same output (and CDV) as the centralized FLFO multiplexer. A compact coding of the timing information restricts bandwidth waste for control information to the size found in other similar protocols. The most important feature is that the scheme is amenable to easy hardware implementation and does not require any call related information.

Performance evaluation of adaptation functions in the ATM environment

The paper models and evaluates key design issues for the Adaptation Layer in ATM networks. The role and efficient design of the adaptation layer is crucial for future B-ISDN based on ATM. We concentrate on packet-mode adaptation services analyzing and simulating relevant protocols employed for narrowband ISDN signaling and packet data. Processing and transmission resources are modeled. In particular we investigate the performance of software based solution resident in micropro- cessors and interfaces to the ATM world via a generic interface component. This set-up which is currently under development is found adequate to handle narrowband ISDN signaling and data traffic. It offers a cost effective and readily available solution for the transition phase towards fully customized broadband components.

A distributed FIFO spacer/multiplexer for access to tree APONs

To bring costs down and make B-ISDN attractive to residential and small businesses from an early stage, cost reductions are necessary. The ATM passive optical networks (APONs) are promising in that respect presenting high reliability, robustness and upgradability. When TDMA (time division multiple access) is employed a medium access control (MAC) protocol is needed to effect bandwidth sharing. However known protocols increase the cell delay variation (CDV) too much over that of an equivalent ordinary ATM multiplexing stage thus producing traffic which creates serious complications to connection acceptance and policing functions of ATM networks. The protocol presented in this paper reduces the CDV to ordinary levels while at the same time operating in a complementary yet decoupled way from the ATM network functions.<<ETX>>

An Integrated Sensor Web Grid Cyberimplementation for Environmental Protection

Wireless Sensor Networks (WSNs) have been deployed for collecting the observed data across different areas. The continuous evolution of wireless networks and sensors emerges a new instrument concept which is called SensorWeb network. This type of network measures data related to geospatial information and can detect the conditions of remote places as a new instrument for environmental monitoring in the physical world. On the other hand the uninterrupted demand of the scientific community for computational and storage resources led us to the Grid Computing. Grid Computing is the combination of several distributed computational resources to a single problem at the same time, usually to a scientific or technical problem that requires a great number of computer processing cycles and a huge storage capability of data.