Vishwanath Ramamurthi

CaDAR: An Efficient Routing Algorithm for a Wireless–Optical Broadband Access Network (WOBAN)

A wireless-optical broadband access network (WOBAN) is a combination of wireless and optical network segments to optimize the cost and performance of an access network. A WOBANs optical backhaul enables it to support high capacity, while its wireless front end [also called a wireless mesh network (WMN)] enables its users to have untethered access. Wireless nodes collect traffic from end users and carry them to the optical part of a WOBAN, possibly using multiple hops, but the traffic also experiences delay at each wireless node. The finite radio capacity at each wireless node limits the capacity on each outgoing link from a wireless node of a WOBAN. Thus, delay and capacity limitation in the WMN of a WOBAN are major constraints. We design a capacity- and delay-aware routing scheme, called CaDAR, to minimize the delay and increase the throughput in the WMN of a WOBAN. Our analysis shows that CaDAR is an efficient routing scheme for WOBAN that can support much higher load and has lower system delay than other approaches (IEEE Network, vol. 22, no. 3, p. 20, 2008) because of better load-balanced routing.

CaDAR: An Efficient Routing Algorithm for Wireless-Optical Broadband Access Network

Hybrid wireless-optical broadband access network (WOBAN) is a combination of wireless and optical networks to optimize the cost and performance of an access network. Wireless nodes collect traffic from end users and carry them to the optical part of a WOBAN using multiple hops, accumulating delay at each wireless node. Moreover, the radio capacity on each wireless link limits the capacity on each outgoing link from the node in a single-radio wireless mesh network (WMN) of a WOBAN. Thus, delay and capacity limitation in the WMN of a WOBAN is a major bottleneck. We design a capacity and delay aware routing scheme, CaDAR, to minimize the delay and increase network support in the WMN of a WOBAN. Our analysis shows that CaDAR is an efficient routing scheme for a single-radio WMN for a WOBAN that can support much higher load and has lower system delay than other approaches because of better load balanced routing.

Link Scheduling and Power Control in Wireless Mesh Networks with Directional Antennas

Directional antennas are very attractive in wireless mesh networks (WMN). We study the problem of link scheduling and power control in a time-division multiple access (TDMA) WMN where the nodes use directional antennas. This is a cross- layer design problem spanning the physical and the link layers. Link scheduling in WMNs requires careful modeling of interference. Interference models used for omni-directional antennas cannot be used for directional antennas. We develop a generalized interference model applicable to directional antennas. Then, we use this model to formulate the link scheduling and power control problem as a Mixed Integer Linear Program. We also propose a heuristic algorithm to solve the problem efficiently.

Channel, capacity, and flow assignment in wireless mesh networks

We study the problem of channel capacity, and flow assignment (CCFA) in multi-channel wireless mesh networks (WMNs). CCFA involves the joint assignment of channels, distribution of wireless capacity, and determination of link flows to enhance the effectiveness of WMNs. We first study the capacity assignment (CA) problem in WMNs (WMN-CA) which involves the distribution of wireless capacity, given the topology and the flows (i.e., traffic demands and routing). Unlike wired networks, the capacities of different wireless links in a WMN have to be carved out of the capacities of wireless nodes. Since the wireless medium is shared by various wireless nodes, interference between different wireless links constrains the distribution of the wireless capacity available at individual nodes. We formulate WMN-CA as a convex non-linear optimization problem (NLP). We also present efficient heuristics to solve the problem and compare their relative performance. We then propose a linear programming (LP) based iterative algorithm for CCFA. We define a new metric, called network utility, which takes both throughput and average packet delay into account and includes a parameter that can be used to emphasize one over the other. Our approach has two separate phases - (1) channel assignment and (2) multi-channel capacity and flow assignment. The two phases of the iterative CCFA algorithm are performed repeatedly to improve the network utility which allows for a tradeoff between delay and throughput that can be achieved in a WMN.

Hybrid Wireless-Optical Broadband Access Network (WOBAN): Capacity Enhancement for Wireless Access

Noting that the optical part of a hybrid wireless- optical broadband access network (WOBAN) has high capacity, we need to enhance the capacity for wireless access using a low-cost solution. Our prior work developed a solution where each wireless node was equipped with a single radio. Deploying multiple radios, say two, at each node will improve the performance of wireless access, but this will also increase the cost of the solution. However, deploying multiple radios at only a few nodes, especially those that are overloaded with traffic, can lead to a less-costly solution, possibly without sacrificing performance. Hence, we study how to optimally place a limited number of additional radios at the wireless nodes to save the overall network cost. We formulate this problem as an Integer Linear Program (ILP) and solve it using a standard solver such as CPLEX. As expected, by deploying multiple radios at bottleneck wireless nodes, we can obtain almost the same performance as a WOBAN with multi-radios at all nodes.

Optimal Capacity Allocation in Wireless Mesh Networks

We address the problem of ldquocapacity allocationrdquo in wireless mesh networks (WMNs). In the ldquocapacity allocationrdquo problem, the network topology (namely the desired wireless links) and routing are given. The problem then is to assign capacities to different links (which are carved out of the radio capacity of a wireless node) in such a way that the average network delay is minimized. This classic problem has been solved by Kleinrock for wired networks subject to a cost constraint. The problem of capacity allocation in wireless networks is different because of the following reasons: i) the capacity of a wireless radio is limited by the physical-layer technology; ii) wireless channel is a shared medium; and iii) the overall capacity of the wireless network is limited by interference. These unique features of a wireless network necessitates a cross-layer approach, involving the physical and the network layers, to solve this problem as opposed to a traditional network planning problem. We use queuing theory along with wireless interference modeling to present a cross-layer solution to this problem.

Enhancing Multi-Hop Wireless Mesh Networks with a Ring Overlay

Wireless mesh network (WMN) has become a popular access network architecture. But because of its multi-hop nature, co-channel interference and contention, the attainable capacity of a wireless node in a WMN is significantly less than the radio capacity. We propose a overlay architecture for a WMN, where a wireless-ring (WRing) is deployed over the regular mesh for carrying wireless mesh traffic only. Our analysis shows that WRing improves the performance of a WMN significantly by reducing the interference and contention.

Green Provisioning of Cloud Services over Wireless-Optical Broadband Access Networks

Todays access networks are increasingly shaped by the services that they provide to the end users. In a hybrid wireless-optical broadband access network (WOBAN), to access any service, connection requests require multi-hop communications over the wireless mesh network (WMN) and the Passive Optical Network (PON), and subsequently over the Internet to some server in the service providers domain. To improve the delivery of such services over WOBAN, we can design a Cloud-Integrated WOBAN (CIW) by deploying a few cloud components (CCs) in the WOBAN itself and serve local services from these local CCs. In this paper, we propose a novel energy-saving routing mechanism, called Green Routing for CIW (GRC), that manages the activation of network components, namely ONUs and CCs, to minimize the overall energy consumption of CIW. It performs load-balanced anycast routing across active devices. Our performance evaluation shows that GRC operates with low average packet delay and achieves significant energy savings by turning off about 50% of the ONUs and about 20% of the CCs.

Cost-efficient design for higher capacity hybrid wireless-optical broadband access network (WOBAN)

Noting that the optical part of a hybrid wireless-optical broadband access network (WOBAN) has high capacity, we need to enhance the capacity for wireless access using a low-cost solution. Deploying multiple radios, say two, at each node will improve the performance of wireless access, but this will also increase the cost and power consumption of the solution. However, deploying multiple radios at only a few nodes, especially those that are overloaded with traffic, can lead to a less-costly and less-power-hungry solution, possibly without sacrificing performance. Hence, we propose a mixed-capacity wireless access (MCWA) architecture for WOBAN and study how to optimally place a limited number of additional radios at the wireless nodes to save the overall network cost. We formulate this problem as a Mixed Integer Linear Program (MILP) and solve it using a standard solver such as CPLEX. An efficient channel and radio assignment scheme is essential to utilize MCWA by reducing interference and contention in the wireless front-end of WOBAN. We propose an intelligent channel and radio assignment (ICRA) for MCWA in WOBAN that performs load balancing over different channels to reduce interference and contention. We formulate this problem as another MILP and solve it using CPLEX. Our findings show that, by using a mixed-capacity wireless access design with intelligent channel and radio assignment, we can obtain almost the same performance as a WOBAN with multi-radios at all nodes.

Cloud-Integrated WOBAN: An offloading-enabled architecture for service-oriented access networks

In a hybrid wireless-optical broadband access network (WOBAN), to access any service, connection requests require multi-hop communications over the wireless mesh network (WMN) and the Passive Optical Network (PON), and subsequently over the Internet to some server in the service provider’s domain. This may cause bottlenecks in the WMN and may result in degraded performance. Today’s access networks are increasingly shaped by the services that they provide to the end users. In this paper, we propose an access network design that integrates a cloud with WOBAN called Cloud-Integrated WOBAN (CIW). CIW creates an infrastructure platform to provide different “cloud services” from within the access network. It has several important benefits: (1) it improves resource utilization by offloading traffic from wireless links, (2) it provides higher scalability by reducing bottleneck from the gateways of WOBAN, and (3) it creates an infrastructure for providers to facilitate different cloud services with their access network. We determine how the cloud components providing the services should be placed in WOBAN in order to optimize the resources while providing better service. We formulate this problem as a Mixed-Integer Linear Program (MILP) and solve it on a realistic study case. We also propose a novel energy-saving routing mechanism, called Green Routing for CIW (GRC), that allows CIW to self-manage the activation of network components, namely ONUs and CCs, to minimize the overall energy consumption of CIW. GRC performs load-balanced anycast routing across active devices. Our performance evaluation shows that GRC achieves significant energy savings with low average packet delay.