Cog-Fi: A cognitive Wi-Fi Channel Hopping architecture for urban MANETs

The success of wireless networks has increased the competition for wireless resources in unlicensed bands by a plethora of wireless devices. Mobile devices that operate in the unlicensed band and are not semi-permanently associated with an access point (either residential or business) find it increasingly difficult to compete in the crowded spectrum. An interesting solution is to equip the mobiles with cognitive radios and allow them to identify the set of least occupied channels and hop across them. In our previous work, we have developed the Cognitive Channel Hopping (CCH) protocol, which utilizes a subset of multiple lightly loaded channels, avoiding channels with heavy external interference. While any routing protocol can work with CCH, conventional protocols like DSR and AODV are not well-suited for CCH networks, as they are not designed to work with multiple channels (in the same MANET) and do not take into account other factors that affect the network performance, such as link rates and channel load. In this work, we define a network architecture Cog-Fi, which exploits CCH over multiple channels in MANETs. We devise a routing protocol called CH-LQSR, which further increases the network performance when used in conjunction with CCH. Our evaluation shows that Cog-Fi can exploit channel diversity without explicit coordination. It outperforms conventional approaches in MANET scenarios with multiple flows and heavy external interference.