Wireless Networks
Wireless Mesh Networking
PI: Samir R. DasÂ
Wireless mesh networks are cost-effective means to provide broadband network service in a geographic area. In this project we will investigate the use of mesh networks for streaming and interactive media applications such as voice and video. We investigate two issues. The first issue concerns fine-grain route adaptation. This is needed as the traditional routing protocols have hysteresis, i.e., they use past information to react to changes in the network and thus are slow to react. The second issue is call admission control for voice-over-IP applications. This admission control uses a capacity model of the network and performs admission control and routing decisions on incoming calls at the egress mesh router. The protocols developed are to be implemented in an indoor mesh networking testbed. (NEC)
Implementation Techniques for Last-Mile Wireless Mesh NetworksÂ
PI: Tzi-cker ChiuehÂ
The explosive growth of IEEE 802.11 wireless LAN (WLAN) technology has prompted a rethinking of how metropolitan-area mobile wireless network services should be designed and implemented. Specifically the cost-effectiveness of 3G cellular networks is being called into question owing to their gargantuan initial licensing cost. In contrast, wireless mesh networks (WMN) that leverage 802.11-based WLAN hardware or the emerging 802.16 wireless broadband access standard (WiMAX) do not incur any licensing fees because they operate in the unlicensed spectrum. Moreover, the hardware/software components of wireless mesh networks are relatively inexpensive since their success in consumer and enterprise market segments creates economies of scale and significantly brings down the manufacturing and development cost. In this proposal, we propose to develop, implement, and evaluate a novel wireless mesh network architecture called Hyacinth, that is specifically designed to support last-mile broadband internet accesses.
Hyacinth features several unique innovations that are not present in other WMNs. First, Hyacinth significantly increases the aggregate throughput of a WMN by supporting multiple WLAN interfaces per WMN node, each of which potentially operates at a distinct radio channel. Second, Hyacinth supports a host-transparent network-layer handoff scheme that preserves continuity of network applications as end hosts move across different subnets, without requiring any modifications to the end hosts. Third, Hyacinth enhances the traffic aggregation capability of each WMN node by equipping each WMN node with an array of access points and dynamically balancing their load. Fourth, the routing protocol used among Hyacinth nodes has built sufficient redundancy into route updates such that compromised nodes that lie can be detected and isolated. We will combine these techniques to build a 16-node wireless mesh network prototype and demonstrate, through a detailed performance study, that the Hyacinth architecture is high-performance, robust, secure, and easy to maintain. (NSF)
Exploiting Spatial and Channel Diversity in Mobile Ad Hoc Networks
PI: Samir R. DasÂ
Mobile ad hoc networks are multi-hop wireless networks, with dynamically changing network topology. Such networks are useful in forming a self-organizing communication network among the first-responders in an emergency rescue scenario or among soldiers and battlefield units in a military operation. Exploiting available network capacity is an important problem in such networks because of the shared nature of the radio medium. One way to exploit capacity is to exploit spatial diversity by using multiple routing paths. In this project we are working on dynamic multipath routing protocols that can achieve a good load balance. The protocols are on-demand and can form link- or node-disjoint, as needed, in a loop-free fashion. The effectiveness of such multipath routes is reduced when paths are formed in the radio vicinity. We are using multiple channels in the underlying medium access control layer so that independent channels can be assigned to neighboring routes. Channel select ion methods for such an approach are important components of our research. In summary, our goal in this project is to exploit spatial and channel diversity in a mobile ad hoc network with a synergistic use of multipath routing and multichannel medium access protocols. (NSF)
Multichannel, Multihop Wireless LAN for Rapid Deployment
PI: Samir R. DasÂ
This project involves an experimental study to design and develop a prototype multihop wireless local area network architecture. The architecture consists of
1) access points that are connected via a backbone wireless network, and
2) client devices that connect to the access points.
We will explore routing protocols for such an architecture, and use of multiple radios in the access points tuned to different bands/channels to provide bandwidth aggregation. The access points can be battery powered. We will study scheduling algorithms so that access points can be powered off if not needed to maintain network coverage. This will improve network lifetime. This project will have an impact on lowering the deployment costs of telecommunication infrastructure and services in "developing" countries such as India and will contribute to reducing the "digital divide" between those who can and cannot use new information and communication tools effectively. This study will also benefit development of wireless local area network architecture that can be deployed rapidly, and thus will be useful for many emergency or tactical operations. (NSF)