Packet switching (Data transmission)

Capture effect has shown considerable improvement on performance of slotted ALOHA systems. Further, improvement is expected by increasing the number of base stations. The performance of such slotted ALOHA systems is analyzed with the aid of Equilibrium point analysis. Packet dropping due to finite number of retransmissions is taken into account. The numerical results indicate that the finite number of retransmission trials mainly contribute to the improvement of the packet dropping probability in the range of light input traffic. The use of multiple base stations improves the overall throughput and the average transmission delay in the range of heavy input traffic.

Multiprocessor systems have demonstrated great potential for meeting the ever increasing demand for higher performance. In this thesis, we develop simulation models with fewer and more realistic assumptions to evaluate the performance of the circuit-switched cluster-based multiprocessor system. We then introduce a packet-switched variation of the cluster-based architecture and develop simulation models to evaluate its performance. The analysis of the cluster-based systems is performed for both uniform and non-uniform memory reference models. We conducted similar analysis for the crossbar and multiple-bus systems. Finally, the results of the cluster-based systems are compared to those obtained for the crossbar and the multiple-bus systems.

Several neural network applications solving practical problems in communications are presented. A neural network algorithm to select paths through a three stage switching network is developed. An analysis of the dynamics of the neural network and a convergence proof are provided. With the help of computer simulations, a four dimensional region for the valid combinations of the neural network parameters was discovered. An analysis is performed to determine the characteristics of this region. The behavior of the neural network algorithm for different switching network configurations and varying traffic patterns were investigated. The effect of initial state of the neural network and heuristic improvements to the algorithm is provided. A comparative analysis of the neural network path selection algorithm against a sequential search method is also given.

A waveform substitution technique using interpolation based on such slow varying parameters of speech as short-time energy and average zero-crossing rate is developed for a packetized speech communication system. The system uses 64 Kbps conventional PCM for encoding and takes advantage of active talkpurts and silence intervals to increase the utilization efficiency of a digital link. The short-time energy and average zero-crossing rates calculated for the purpose of determining talkpurts are transmitted in a preceeding packet. Hence, when a packet is pronounced "lost", its envelope and frequency characteristics are obtained from the previous packet and used to synthetize a substitution waveform which is free of annoying sounds that are due to abrupt changes in amplitude. Informal listening tests show that tolerable packet loss rate up to 40% are achievable with these procedures.

Rapid advances in technology have resulted in the evolution of Integrated Services Digital
Networks (ISDNs) to Broadband ISDNs (BISDNs). This thesis discusses and evaluates the
performance of a high speed and high capacity packet switching system architecture for
BISDNs. This system supports various high speed communication services like data, voice,
and video services. The main performance criterion is the mean switching delay, which
is defined as the packet transfer delay through the bus and through the loop connecting
various switch modules in the system. Other performance criteria are power and system
throughput. The performance parameters are evaluated separately for both data and control
(signaling) packets. The effects of the number of switch modules in the system, the data
packet length, and the ratio of arrival rate of data to control packets are examined.
It is observed that the switching delay and throughput increase with an increase in the
traffic intensity for any number of switching modules and the ratio of arrival rate of data
to control packets. The delay for data packets is found to be higher for lower values of
this ratio and vice versa. The power is found to be maximum for about 70% load. The
switching delay in this simulation model is found to be less than one millisecond for data
packets.

Ad-hoc wireless networks are the networks without an infrastructure. Infrastructure-less mobile networks have no fixed routers or base stations. All nodes in these networks are capable of movement and can be connected dynamically at any time. The issue of routing packets effectively between any nodes in an ad-hoc network is a challenging task. A path that was considered optimal at a given time point may not be working after a few moments. Moreover the bandwidth available between a link may not remain the same since every node in this kind of network is behaving as a router and may experience drastical increase in throughput and traffic load. For the purpose of quality of communication, bandwidth reservation within ad-hoc wireless network nodes may prove promising as a solution to this problem. This thesis presents an approach to reserve bandwidth available utilizing CDMA of bandwidth to the Ad-hoc Distance Vector Protocol for the ad-hoc wireless networks. Various quality of service issues are discussed and are tested for optimal performance for AODV routing protocol. A QoS performance comparison based on mobility, congestion and throughput is made between AODV and CDMA based AODV mobile routing protocols.

The objective of this research is to determine the macroscopic behavior of packet transit-times across the global Internet cloud using an artificial neural network (ANN). Specifically, the problem addressed here refers to using a "fast-convergent" ANN for the purpose indicated. The underlying principle of fast-convergence is that, the data presented in training and prediction modes of the ANN is in the entropy (information-theoretic) domain, and the associated annealing process is "tuned" to adopt only the useful information content and discard the posentropy part of the data presented. To demonstrate the efficacy of the research pursued, a feedforward ANN structure is developed and the necessary transformations required to convert the input data from the parametric-domain to the entropy-domain (and a corresponding inverse transformation) are followed so as to retrieve the output in parametric-domain. The fast-convergent or fast-computing ANN (FC-ANN) developed is deployed to predict the packet-transit performance across the Internet. (Abstract shortened by UMI.)

This thesis is concerned with the performance analysis of a DS/CDMA packet radio system under Rician fading channel conditions. Analytic expressions are derived for the probability density functions and cumulative distribution functions of the total signal-to-interference ratio when considering mixed-type of interference sources. The users are divided into two groups: those that apply closed-loop power control and those that apply open-loop power control. Four different scenarios of power control were studied. The outage probability is evaluated for the four different cases of power control. A packet transmission protocol with forward error correction capabilities is considered, and the probability density functions of the signal-to-interference ratio are used to calculate the average block error probability as well as the outage probability of the system.

This thesis proposes a design for a signaling gateway to address the issues of interoperability between H.323 and SIP. The two IP Telephony standards currently compete for the dominance of IP telephony protocols. We investigate and study these two protocols in terms of interoperability. The question central to this thesis is the issue of interoperability between H.323 and SIP. A signaling gateway model is proposed and designed to address and solve this issue. The proposed model includes Call Initialization, Call Setup and Control, Capability Exchange and Call Termination, which are the four fundamental features supported by H.323 and SIP for establishing calls. Furthermore, we design the four internal components, which are common to the four main functions mentioned above. These four components are type checking, decomposition, conversion and reformatting. In addition, we illustrate in six different cases the functionality of the proposed signaling gateway for establishing calls between H.323 and SIP. These six cases also demonstrate the conversion capability of the proposed signaling gateway during a call between H.323 and SIP endpoints.

The research reported in this dissertation studies ATM multiplexing and connection admission control schemes for traffic management in connection-oriented packet networks. A new threshold-based connection admission control scheme is proposed and analyzed. The scheme uses effective bandwidth to make decision whether to accept or reject the connection request. This threshold specified effective-bandwidth method is first simulated on a simple 4-node connection-oriented packet network model, and then extended to a more complex 8-node network model under a variety of environments. To reduce the cell-loss ratio when the arrival rates of the connection requests are large, the dynamic effective bandwidth mechanism is proposed and relevant simulations are addressed on the two network models. The traffic used in the simulation is a multiplexed stream of cells from video, voice and data sources, which is typical in ATM environments. The multiplexed traffic is generated using a discrete event scheduling method. The simulation programs for the 4-node network model and for the 8-node network model are verified by the theoretical values of the blocking probabilities of the connection requests, and Little's Theorem. Simulations on the two network models show similar results. Pertinent to a network that supplying several service categories, the threshold-based connection admission control is shown to affect the blocking probabilities of each type of traffic. In some environments, having a threshold is advantageous over the case without a threshold in terms of cell-loss ratio, cell transfer delay and power (throughput divided by cell transfer delay). The simulation results also show that the dynamic effective bandwidth based method helps to reduce the cell-loss ratio significantly when the arrival rates of the connection requests are large.