ABSTRACT The number of cellular communication subscribers continues to grow, attesting to the great success of this technology. However, cellular networks have inherent limitations on cell capacity and coverage and shortcomings such as the dead spot and the hot spot problems. Multi‐hop cellular networks (MCNs) help enhance the cell capacity and coverage, while, at the same time, alleviating the dead spot and hot spot problems, increasing the utilization of radio resource, and reducing the power consumption of mobile terminals. In the past decade, more than a dozen of MCN architectures were proposed. In this paper, we study various types of MCN proposals. We identify and discuss the design decision factors and use these factors to classify most existing MCN proposals. Future research directions, including studies of capacity and energy consumption, and approaches addressing design issues such as cell size, routing, channel assignment, load balancing for MCNs are discussed. Copyright © 2010 John Wiley & Sons, Ltd.

Multihop cellular networks (MCNs) enhance the capacity and coverage and alleviate the dead-spot and hot-spot problems of cellular networks. They also allow faster and cheaper deployment of cellular networks. A fundamental issue of these networks is packet delay because multihop relaying for signals is involved. An effective channel assignment is the key to reducing delay. In this paper, we propose an optimal and a heuristic channel assignment scheme, called OCA and minimum slot waiting first (MSWF), respectively, for a time division duplex (TDD) wideband code division multiple access (W-CDMA) MCN. OCA provides an optimal solution in minimizing packet delay and can be used as an unbiased or benchmark tool for comparison among different network conditions or networking schemes. However, OCA is computationally expensive and, thus, inefficient for large real-time channel assignment problem. In this case, MSWF is more appropriate. Simulation results show that MSWF achieves on average 95 percent of the delay performance of OCA and is effective in achieving high throughput and low packet delay in conditions of different cell sizes.

In 3G-based multihop cellular networks (MCNs), the cell size affects the cell capacity and the network reachability, which in turn affect the total demand of source nodes that can be served or the system throughput. Improper cell size assignment greatly affects the performance of the networks. To address the cell size issue, we recently proposed the optimal cell size (OCS) scheme to find optimal cell sizes to maximize the system throughput for a 3G TDD W-CDMA MCN. Although OCS provides an optimal cell size solution, it is computationally expensive. In this paper, we propose a heuristic cell size scheme, called Small Cell Size First (SCSF), which is more efficient and provides good results in terms of throughput compared to the optimal solutions provided by OCS. SCSF outperforms the fixed small cell size (SCS) multi-hop case when the network is sparse and the large cell size single-hop case regardless of the network density.

3G-based Multi-hop cellular networks (MCNs) inherit a special characteristic of 3G systems, namely the relationship between a cell's coverage and its capacity. At the planning stage, a cell may be statically set to have small coverage with high capacity, a large coverage with small capacity, or some other fixed settings in between. Such static settings, however, do not adapt to the projected dynamic nature of users in 3G-especially in an MCN environment. In this paper, we propose the Optimal Cell Size (OCS) scheme for a 3G TDD W-CDMA MCN multi-cell environment. Given the cell capacity function, users' distribution and demands, OCS yields optimal cell sizes that maximize the system throughput through balancing coverage and capacity. Not only can OCS cope with dynamic network conditions, but it can also be considered as an aid to the network planning process. To the best of our knowledge, this is the first multi-cell optimal cell size scheme in the context of MCNs.

Wireless networks have made great gains in usability and popularity. However, inherent limitations on cell capacity and coverage still exist. There are also dead-spots and hotspots problems in these networks. Ad hoc multi-hop relaying enhances cell capacity and coverage, alleviates the dead-spots problem, and helps to ease congestion in hotspots. However, multi-hopping also increases packet delay. Effective channel assignment is key to reducing delay. Existing channel assignment schemes are heuristics and may not guarantee optimal solutions in terms of minimum delay. In this paper, we provide an optimal channel assignment (OCA) scheme for ad hoc TDD W-CDMA multi-hop cellular networks (MCN) to minimize packet delay. OCA can also be used as an un-biased tool for the comparison among different network topologies, network densities, and protocols. To the best of our knowledge, this is the first time that a minimum delay optimal channel assignment is proposed in a multi-hop cellular environment.

Multi-hop relaying is an important concept in tackling the inherent problems of limited capacity and coverage in cellular networks. It helps to solve the dead-spots problem and to ease congestion in hotspots. However, to obtain good performance of multi-hop relaying, an effective channel assignment scheme is needed. In this paper, we study the design goals of a good channel assignment scheme. We then propose a channel assignment scheme, called Extended Delay-Sensitive Slot Assignment (E-DSSA), which achieves all these goals. The distinctive feature of E-DSSA is the use of a novel transmission zone testing technique which allows high flexibility and precision in channel assignment in TDD W-CDMA multi-hop cellular environment. Performance evaluation shows that E-DSSA outperforms its existing counterparts in terms of data throughput with low delay for both sparse and dense networks. E-DSSA is also shown to adapt to different cell sizes achieving high throughput and call acceptance ratios.

Multi-hop relaying is an important concept in future generation wireless networks. It can address the inherent problems of limited capacity and coverage in cellular networks. However, most multi-hop relaying architectures are designed based on a small fixed-cell-size and a dense network. In a sparse network, the throughput and call acceptance ratio degrades because distant mobile nodes cannot reach the base station to use the available capacity. In addition, a fixed-cell-size cannot adapt to the dynamic changes of traffic pattern and network topology. In this paper, we propose a novel multi-hop relaying architecture called the adaptive multi-hop cellular architecture (AMC). AMC adapts the cell size to an optimal value that maximizes throughput by taking into account the dynamic changes of network density, traffic patterns, and network topology. To the best of our knowledge, this is the first time that adaptive (or optimal) cell size is accounted for in a multi-hop cellular environment. AMC also achieves the design goals of a good multi-hop relaying architecture. Simulation results show that AMC outperforms a fixed-cell-size multi-hop cellular architecture and a single-hop case in terms of data throughput, and call acceptance ratio