Abstract

As WiFi becomes a universal means for wireless access to Internet, the multi-cell WLAN with a lot of access points (APs) deployed densely is not seldom but comes to be a typical configuration of WLANs, especially, in the hot-spots such as shopping malls, air ports, universities, and enterprises. Moreover, in the large classrooms or conference rooms, a huge volume of multimedia traffic should be multicasted to the users via a number of APs. Since more than one AP can serve a user in a multi-cell WLAN, the overall WLAN performance highly depends on the user association (UA) scheme that decides the serving AP for each user. On the other hand, since the traffic load in such a multi-cell WLAN varies largely with time (e.g., day-time and night), the sleep mode operation of AP can be very helpful for reducing the energy consumption in the entire system. There have been some UA schemes, each of which takes account of only the load balancing between the APs, only the multicast transmission, or just the sleep mode operation of APs. However, to maximize the performance of a multi-cell WLAN, these three problems should be considered together. In this paper, we propose a UA scheme which simultaneously addresses the load balancing, the multicast transmission, and the AP sleep operation adapting to traffic load. The proposed UA scheme is formulated into a mixed integer linear programming problem which can be optimally solved by using branch and bound technique. We implement the proposed scheme using real devices and compare its performance with those of the existing UA schemes. The experimental results show that the proposed scheme greatly outperforms the existing schemes, in terms of the throughput per working AP, the number of working APs, and a load balancing factor.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call