The commercial success of cellular networks, combined with advances in digital electronics, signal processing, and telecommunications research have lead to the design of next generation wireless systems. The next generation wireless system (NGWS) is a system of globally available, wireless multimedia networks that are anticipated to deliver service to a mobile computer terminal "anywhere at anytime." In this dissertation, new handoff techniques were developed to support global roaming with quality of service constraints within NG wireless systems. Whereas conventional handoff techniques support single connection terminals that operate within a homogeneous network, NG wireless systems promise to support terminals with multiple connections carrying different types of traffic, with varying quality of service constraints, which may handoff between different tiers of the same network, or between different types of networks. For intra-system roaming, a new handoff technique for real-time traffic in Mobile IP version 6 networks was created to adapt to IP-based quality of service architectures. Disruption of the communication path, bandwidth expenses, and buffering requirements were reduced. Next, a new handoff technique was developed for Wireless ATM networks that used the source switch to manage connections with multiple mobile endpoints, and to reroute connections according to the type of traffic being carried. For inter-system roaming, new boundary elements were introduced to the NGWS architecture. A new inter-system handoff signaling and rerouting protocol was created to enable format transformations and advanced preparation for mobile terminals that may roam between a variety of networks. Finally, an admission control algorithm for inter-system roaming was created to provide a mechanism for quality of service re-negotiation and to regulate the admission non-subscriber traffic.

The full dissertation is available here
Ad hoc networks are gaining increasing popularity in recent years because of their ease of deployment. No wired base station or infrastructure is supported, and each host communicasts one another via packet radios. In ad hoc networks, routing protocols are challenged with establishing and maintaining multihop routes in the face of mobility, bandwidth limitation and power constraints. In this dissertation, we study the routing strategies for ad hoc networks. On-demand routing protocols and table-driven algorithms are analyzed and compared against each other. Our study shows that on-demand protocols are better suited for mobile networks because they generate less control overhead and manage the mobility in a more efficient manner. Simulation experiments also indicate that providing multiple routes is beneficial in increasing the robustness against mobility.

We investigate the scalability characteristics of on-demand routing protocols and propose schemes to enhance the performance. We also study the interaction between MAC (Medium Access Control) and routing protocols by simulation.

Based on the lessons learned from the performance evaluation studies, we design new on-demand protocols. We introduce three unicast routing algorithms with different approaches. AODV-BR (Ad hoc On-demand Distance Vector with Backup Routes) is a scheme applied to the existing AODV protocol for establishing backup routes while the primary route is constructed, without transmitting additional control messages. Backup routes are utilized when the primary path is disconnected. The Split Multipath Routing (SMR) protocol builds maximally disjoint routes. Providing multiple routes helps minimizing route recovery process and reducing control overhead. Distributing traffic into multipaths prevents nodes from being congested. Dynamic Load Aware Routing (DLAR) is a protocol that uses the load of the intermediate nodes instead of the shortest distance, as the main route selection metric. The protocol attempts to avoid building routes with congested links.

We then present the On-Demand Multicast Routing Protocol (ODMRP), a novel multicasting scheme that utilizes a mesh structure. Multiple routes created by the mesh make the protocol robust to mobility. Multicast routes and group membership are obtained on demand to use the network resources efficiently and effectively. Simulation study shows that ODMRP outperforms other popular multicast protocols.

Research in Nomadic Computing has surfaced, in the early 90s, with the emergence of a new computing platform, pushed by technology improvements. This novel research area, addressed here in its Distributed Systems aspect, has gathered concepts and theory from other classic areas, such as Databases, Communications and Collaborative Systems.

Nomadic computing introduces several evolution paths over prior models of operation. A clear issue is the need to adapt replication mechanisms, and the identification of the limitations in centralized replication control schemes. Long periods of isolated operation, and user centered data (mobile devices are rarely shared among users), discourages the use of pessimistic systems and points to the adoption of optimism and concurrent evolution. In fact, some of these tendencies are shared with the neighbour area of Large Scale Distributed Systems, where similar conditions are present.

The study carried out in this thesis is focused on Autonomous Replication and a means for Autonomous Operation. This approach attempts to meet the needs of the new operation patterns and to supply a theoretical context that helps constructing replicated systems that are independent from a centralized management of replica affiliation. Another target is to identify mechanisms of information sharing, between replicated entities, that allow some degree of control over the evolution of shared state.

Another key contribution of this thesis is on mechanisms for capturing causality in an arbitrary number of replicas that have evolved independently. Whilst motivated by mobility, these theoretical tools find a broader application area on classical distributed systems.

A final element is the characterization of abstract data types that, once subject to arbitrary replication with optimistic operation, still provide reconciliation guarantees. This work is needed to identify which data types allow automatic reconciliation policies in the presence of symmetrical replicas.

The full dissertation is available here
Telecommunication service providers have recently begun to offer ubiquitous access to packetised data. As a result, the Internet is not limited to computers that are physically connected but is also available to users that are equipped with mobile devices. This ubiquitous access fuels the growth and the usage of the Internet even further, and thus the realisation of dynamic Internet. With the realisation of the dynamic Internet, increasing support is needed for Internet protocol (IP) and transmission control protocol (TCP) over wireless/mobile networks.

Two areas of interest in this thesis are unicast and multicast routing in connection-less and connection-oriented networks. To address the problems of routing protocols in mobile computing environments, the active networks (ANs) paradigm is employed. ANs provide an alternative paradigm to solving network problems and comprise programmable network elements that allow enhancement of existing protocols and the execution of active protocols which run for the duration of the communication session.

This thesis investigates the viability and advantages of ANs when applied to routing in mobile communications. Two new AN-based protocols, for IP and asynchronous transfer mode (ATM) networks, that address the problems of multicast routing with mobile group members are outlined. The Internet Engineering Task Force (IETF) mobile IP has been augmented with active programs in order to enhance its operation further. Also, a novel model for rerouting connections in ATM networks is presented.

Results of extensive simulation studies comparing performances of conventional as well as some recently proposed protocols with those of AN-based protocols are presented. The results obtained from these simulation studies show that AN-based protocols have the following benefits: (i) efficient adaptation to mobility, (ii) reduced signaling overheads, (iii) high reuse of allocated network states, (iv) extensibility, (v) network topology independence, and (vi) scalability. The aforementioned points are crucial in mobile environments where states at routers (switches) are frequently updated due to mobility. It was shown that ANs provide the most benefits to protocols that maintain states within the network, for example connection-oriented and multicast protocols. AN-based protocols enable fast and efficient update of the states maintained at the routers/switches without incurring excessive signaling overheads. Moreover, part of a connection or multicast tree can be updated iteratively with the use of ANs, resulting only in modifications to routers (switches) that are affected by host migration. A model for deploying active programs that is coupled with the the protocol operation is also demonstrated. Implementation of such a model eliminates the need for strategic positioning of active services.

The full dissertation is available here
Wireless communications systems of the future will experience more dynamic channel conditions and a wider range of application requirements than systems of today. Such systems will require exible signal processing algorithms that can exploit wireless channel conditions and knowledge of end-to-end user requirements to provide effcient communications services.

In this thesis, we investigate wireless communications systems in which the signal processing algorithms are specifically designed to provide efficient and flexible end-to-end functionality. We describe a design approach for exible algorithms that begins with the identification of specific modes of exibility that enable efficient over-all system operation. The approach then uses explicit knowledge of the relationships between the input and output samples to develop efficient algorithms that provide the desired exible behavior. Using this approach, we have designed a suite of novel algorithms for essential physical layer functions. These algorithms provide both dynamic functionality and efficient computational performance.

We present a new technique that directly synthesizes digital waveforms from pre-computed samples, a matched filter detector that uses multiple threshold tests to provide efficient and controlled performance under variable noise conditions, and a novel approach to narrowband channel filtering. The computational complexity of the filtering algorithm depends only on the output sample rate and the level of interference present in the wideband input signal. This is contrast to conventional approaches, where the complexity depends upon the input sample rate. This is achieved using a composite digital filter that performs efficient frequency translation and a technique to control the channel filter output quality while reducing its computational requirements through random sub-sampling.

Finally, we describe an implementation of these algorithms in a software radio system as part of the SpectrumWare Project at MIT.

The full dissertation is available here
Wireless communications systems of the future will experience more dynamic channel conditions and a wider range of application requirements than systems of today. Such systems will require exible signal processing algorithms that can exploit wireless channel conditions and knowledge of end-to-end user requirements to provide effcient communications services.

In this thesis, we investigate wireless communications systems in which the signal processing algorithms are specifically designed to provide efficient and flexible end-to-end functionality. We describe a design approach for exible algorithms that begins with the identification of specific modes of exibility that enable efficient over-all system operation. The approach then uses explicit knowledge of the relationships between the input and output samples to develop efficient algorithms that provide the desired exible behavior. Using this approach, we have designed a suite of novel algorithms for essential physical layer functions. These algorithms provide both dynamic functionality and efficient computational performance.

We present a new technique that directly synthesizes digital waveforms from pre-computed samples, a matched filter detector that uses multiple threshold tests to provide efficient and controlled performance under variable noise conditions, and a novel approach to narrowband channel filtering. The computational complexity of the filtering algorithm depends only on the output sample rate and the level of interference present in the wideband input signal. This is contrast to conventional approaches, where the complexity depends upon the input sample rate. This is achieved using a composite digital filter that performs efficient frequency translation and a technique to control the channel filter output quality while reducing its computational requirements through random sub-sampling.

Finally, we describe an implementation of these algorithms in a software radio system as part of the SpectrumWare Project at MIT.