Multihop Cellular Networks (MCNs) form combined wireless paradigm that carries the advantages of both traditional cellular networks and wireless multihop relay. Cellular networks depend on a fixed infrastructure to provide wide area coverage for users with high mobility profile. Multihop relay networks depend on wireless devices inside the network to relay signals through multiple hops from source to destination. MCNs were proposed to overcome inherent drawbacks in cellular networks like congestion and dead spots. These gains build on the characteristics of multihop relay that result in increased capacity, decrease energy depletion and virtually extended coverage. But while these gains have been widely accepted and advocated, they have not been verified in rigor. A realistic need therefore exists to quantify these gains in order to realize more capable network management functionalities for this new paradigm. In this thesis, we present an analytical framework for MCNs. We quantify the capacity and energy consumption in MCNs, while considering various call distributions, network loads and transmission power. We apply our framework to Code Division Multiple Access (CDMA) cellular networks, which are very dependent on interference levels in their performance. Our results show that capacity can be increased in CDMA cellular networks using multihop relay by increasing either the number of simultaneous calls or data rates. We also demonstrate that consumed energy is decreased in MCNs, especially in environments with high path loss. We validate that multihop relay is most rewarding when calls tend to originate near cell borders. Beyond verifying basic claims, we explore other potential gains of MCNs. We investigate the viability of congestion relief and load balancing and substantiate the benefits for congested cells neighbored by lightly loaded cells. Load balancing has also been shown to increase data rates and fairness in user allocations. Lastly, we explore enabling multimedia applications in MCNs and study the application of data rate adaptations given multiple classes of service. A key advantage of our work is that, while applied to CDMA in this thesis, the presented analytical framework can be extended to other technologies. The framework also accommodates both mobile and fixed network relay elements, expanding its applicability to next generation cellular networks. / Thesis (Ph.D, Electrical & Computer Engineering) -- Queen's University, 2009-01-30 09:34:39.735

In a multihop cellular network, mobile terminals have the capability to transmit directly to other mobile terminals enabling them to use other terminals as relays to forward traffic towards the base station. Previous studies of networks consisting of a single cell found that the SINR in a multihop cellular network is slightly lower than in a traditional cellular network. However, multihop cellular networks may have a higher capacity than traditional cellular networks due to their potential for lower intercell interference. For this reason, the effects of intercell interference are investigated in this thesis. Our simulations of a network with many cells show that multihop cellular networks have a higher SINR than traditional cellular networks due to the near elimination of intercell interference. However, multihop cellular networks still suffer from large amounts of interference surrounding the base station because all traffic either emanates or is destined to the base station making it the capacity bottleneck. To resolve this problem, we propose a novel architecture called the autonomous infrastructure multihop cellular network where users can connect their mobile terminals to the backbone network giving them the functionality of an access point. Access points receive traffic from other terminals and send it directly onto the backbone, as would a base station. This reduces the traffic handled by the base station and increases network capacity. Our analysis and simulations show that in autonomous infrastructure multihop cellular networks, the SINR at the base station is higher, the power consumption is lower and the coverage is better than in normal multihop cellular networks. Access points require parameters like their transmission range to be adjusted autonomously to optimal levels. In this thesis, we propose an autonomous pilot power protocol. Our results show that by adjusting a parameter within the protocol, a required coverage level of terminals can be specified and achieved without knowledge of the location or density of mobile terminals. Furthermore, we show that the protocol determines the transmission range that is optimal in terms of SINR and power consumption that achieves the required coverage while effectively eliminating the bottleneck that existed at the base station.

In a multihop cellular network, mobile terminals have the capability to transmit directly to other mobile terminals enabling them to use other terminals as relays to forward traffic towards the base station. Previous studies of networks consisting of a single cell found that the SINR in a multihop cellular network is slightly lower than in a traditional cellular network. However, multihop cellular networks may have a higher capacity than traditional cellular networks due to their potential for lower intercell interference. For this reason, the effects of intercell interference are investigated in this thesis. Our simulations of a network with many cells show that multihop cellular networks have a higher SINR than traditional cellular networks due to the near elimination of intercell interference. However, multihop cellular networks still suffer from large amounts of interference surrounding the base station because all traffic either emanates or is destined to the base station making it the capacity bottleneck. To resolve this problem, we propose a novel architecture called the autonomous infrastructure multihop cellular network where users can connect their mobile terminals to the backbone network giving them the functionality of an access point. Access points receive traffic from other terminals and send it directly onto the backbone, as would a base station. This reduces the traffic handled by the base station and increases network capacity. Our analysis and simulations show that in autonomous infrastructure multihop cellular networks, the SINR at the base station is higher, the power consumption is lower and the coverage is better than in normal multihop cellular networks. Access points require parameters like their transmission range to be adjusted autonomously to optimal levels. In this thesis, we propose an autonomous pilot power protocol. Our results show that by adjusting a parameter within the protocol, a required coverage level of terminals can be specified and achieved without knowledge of the location or density of mobile terminals. Furthermore, we show that the protocol determines the transmission range that is optimal in terms of SINR and power consumption that achieves the required coverage while effectively eliminating the bottleneck that existed at the base station.