Power Minimization and Optimum ONU Placements in Integrated Wireless Optical Access Networks

The deployment of optical fibre in place of copper cable in access networks has experienced remarkable growth over the past several years due to a wide range of benefits. A major benefit of optical fibre over copper cable is that it is more secure and immune to electromagnetic interferences. Optical fibre has also provided the capability of handling higher throughputs for longer distances, and experiences no crosstalk between other fibre optic cables. However, the last mile reach to end-users with optical fibre is very costly. This alternative replacement results in increased costs for manual labour and energy consumption in the access network. The current demand in all areas of telecommunications, and especially access networks, is greener networking. In order to offset the high costs of optical access implementations and to satisfy this demand, an investigation into integrated wireless optical access networks (IWOAN) is warranted.;The proliferation of wireless devices has also motivated the interest in IWOAN as it combines the flexibility and efficiency of wireless with the security and stability provided by optical. With the emergence of smart phones and tablets, wireless access networks are now supporting an increasing amount of traffic volume with improved throughput and accessibility. We employ a Passive Optical Network (PON) infrastructure from the central office to the customer, traced from the Optical Line Terminal (OLT) to the customer premises devices known as Optical Network Units (ONUs) for IWOAN. At the ONU, the optical fibre is terminated and wireless communication is implemented. The ONU acts as a wireless access point/gateway for wireless Base Stations (BS) serving different coverage areas in point-to-point topology. With recent trends of advanced wireless technologies, premium rich applications such as multimedia streaming, interactive gaming and cloud computing are delivered in a satisfactory and economic way. This wireless-optical integration aims to reduce and solve the cost of replacing copper cables. However, another issue is raised with increased costs in energy consumption due to the integration of wireless and optical communication. Typically a large number of ONUs need to be deployed in order to serve many wireless BSs located in different coverage areas. As a result, any cost savings gained by the integration process is exhausted with the increased cost of power consumption.;Energy efficiency is a vital issue to solve. The trade-off for reducing the cost of optical implementations is increased cost of energy consumption. Hence, this thesis proposes an optimized approach to IWOAN through the means of power minimization. We formulate an algorithm that will design the network layout for optimum placements of ONUs to support the BS demands, while minimizing power consumption. We have developed two formulations to measure minimum power consumption. Initially we formulate power consumption on the foundation of aggregate uplink traffic rates at the ONU and the distance for transmission between an ONU and its BSs. We then modify this formulation as a function of uplink traffic rates at the ONU, exclusively. By modeling two formulations we achieve an important observation; the total power consumption is primarily dependent on the transmission power and less on traffic power. We have developed several in-house simulation programs of the network using CPLEX optimization studio. We evaluate our original formulation, and then evaluate the modified formulation and compare results to obtain various power components. We obtain three power components: start-up power, and two dynamic powers; traffic power and transmission power, which all summate to our minimum power consumption. Constraints are used to create an ideal network with system limitations and acceptability. These include channel assignment, ONU installation, network capacity and signal quality.;Our in-house simulation program measures several effects on power consumption: effects of initial placements of ONUs; ONU transmission distances; increase in the total number of BSs in the network, and; channel reuse method using cell structures. In all scenarios we map out optimum placements of ONUs, determine the minimum number of active ONUs, and the minimum power consumption.