Shared Path Protection Algorithms In Hybrid Wireless-Optical Broadband Access Networks

Hybrid Wireless-Optical Broadband Access Network (HOWBAN) is a promising architecture for future access networks, since it combines the advantages of optical access networks and wireless access networks. It shows attractive features such as large transmission capacity, wide coverage, and flexibility. While the optical backhaul of HOWBAN can support higher data rates, it is vulnerable to failure due to its tree architecture, which may lead to huge loss of data. Consequently, effective survivability mechanisms are necessary to minimize the data loss in HOWBAN.This paper first reviews the network structure and the research status of HOWBAN, mainly focusing on the key issues such as energy saving, deployment strategy, routing of HOWBAN. Secondly, this paper investigates the survivability problem of HOWBAN, especially studies the protection mechanism by exploiting the properties of wireless connectivity of the front-end networks to reroute the affected traffic of the back-end networks. Moreover, with a consideration of different bandwidth and availability requirements for different users at different time, we propose a User Behavior Aware Shared Path Protection Algorithm (UBA-SPP) to choose backup optical network unit (ONU) and allocate backup resources. This algorithm ensures the requirement of users in high availability period by using low load ONU to provision backup resources, which is helpful to achieve load balance for ONUs and to reduce the occupation of backup resources as much as possible. In addition, since different users may require different protection bandwidth under network failures, bandwidth degraded-service can be considered in HOWBAN. We propose a Degraded Service Shared Path Protection Algorithm (DS-SPP) to further save the backup resources. With a comprehensive consideration of connection availability, bandwidth demand and degraded-service for different network users, DS-SPP chooses backup ONUs with less reservation of backup bandwidth. And then, it reroutes the affected traffic from the work ONU to the backup ONU through the front wireless mesh network (WMN). Finally, based on VC++6.0 software, a HOWBAN simulation platform is developed to evaluate the performance of our algorithms and to compare them with the existed algorithms proposed in the previous literatures. The simulation results show that UBA-SPP and DS-SPP algorithms can not only guarantee the users’requirements, but also reduce the reserved backup resources and to improve the nework resource utilization.