Resource Allocation for QoS Support in Wireless Mesh Networks

Resource Allocation for QoS Support in Wireless Mesh Networks


Many next generation applications (such as video flows) are likely to have associated minimum data rate requirements in order to ensure satisfactory quality as perceived by end-users. In this paper, we develop a framework to address the problem of maximizing the aggregate utility of traffic flows in a multi-hop wireless network, with constraints imposed both due to self-interference and minimum rate requirements. The parameters that are tuned in order to maximize the utility are (i) transmission powers of individual nodes and (ii) the channels assigned to the different communication links. Our framework is based on using across-decomposition technique that takes both inter-flow interference and self-interference into account. The output of our framework is a schedule that dictates what links are to be activated in each slot and the parameters associated with each of those links. If the minimum rate constraint cannot be satisfied for all of the flows, the framework intelligently rejects a sub-set of the flows and recomputes a schedule for the remaining flows. We also design an admission control module that determines if new flows can be admitted without violating the rate requirements of the existing flows in the network. We provide numerical results to demonstrate the efficacy of our framework.

PROJECT OUTPUT VIDEO: (Click the below link to see the project output video):


The problem of resource allocation and congestion control in wired networks has received a lot of attention. In their seminal work, Kelly et al. have modeled the problem of flow control as an optimization problem where the objective is to maximize the aggregate utility of elastic traffic sources subject to capacity constraints on the links that compose the network. Inspired by Kelly’s work, there has been follow up work, where TCP congestion control is modeled a convex optimization problem, the objective being the maximization of an aggregate user utility; in these efforts distributed primaldual solutions to the problem are proposed.


In contrast with wireline networks, the capacity of a wireless link is not dependent on other flows in the network but on other flows that use links on the same channel (and that are close enough) and external interference. The dependencies between flows is regulated by the protocols at both the link and transport layers. However, these prior efforts do not consider the provision of quality-of-service in terms of supporting minimum rates to the flows that share the network. More importantly, the QoS needs to be provided under conditions of self-interference, where the packets of a flow interfere with other packets that belong to the same flow along a multi-hop path.


In this paper, we propose a framework for maximizing the aggregate utility of traffic sources while adhering to the capacity constraints of each link and the minimum rate requirements imposed by each of the sources. The framework takes into account the self-interference of flows and assigns (a) channels (b) transmission power levels and (c) time slots to each link such that the above objective is achieved. It dictates the rates at which each traffic source will send packets such that the minimum rate requirements of all coexisting flows are met. If the minimum rate requirements of all the flows cannot be met, the framework rejects a subset of flows (based on fairness considerations) and recomputes the schedule and allocates resources to each of the remaining flows.


  • The framework maximizes the aggregate utility of flows taking into account constraints that arise due to self-interference (wireless channel imposed constraints) and minimum rate requirements of sources (QoS requirements).
  • If a solution is not feasible, the framework selectively drops a few of the sources and redistributes the resources among the others in a way that their QoS requirements are met.
  • The proposed framework readily leads to a simple and effective admission control mechanism.
  • We demonstrate the efficacy of our approach with numerical results. We also theoretically compute performance bounds with our network, as compared with an optimal strategy.



  • Creating System Model
  • Channel Assignment
  • Resource allocation
  • Admission control module


Creating System Model

We consider a pre-planned WMN consisting of a set of stationary wireless nodes (routers) connected by a set L of unidirectional links. Some of the nodes are assumed to have the ability to perform functions of the gateway, and one of them is selected to act as the gateway to the Internet. Each node is equipped with a single network interface card (NIC) and is associated with one of C orthogonal (non-overlapping) channels for transmitting or receiving. A sender-receiver pair can communicate with each other only if both of them are tuned to the same channel. In this work dynamic channel switching is assumed to be possible with the NIC. Nodes operate in a half-duplex manner so that at any given time a node can either transmit or receive (but not both). In addition, it is assumed that the network operates in a time-slotted mode; time is divided into slots of equal duration.

Channel Assignment

The proposed algorithm allocates channels in a way that (a) self-interference is avoided and (b) co-channel interference levels among links that use the same channel are kept as low as possible. With our algorithm, links with higher costs are assigned higher priorities in terms of channel assignment over the links with lower cost. This is because links with higher costs suffer from higher levels of congestion and thus, scheduling these links is harder. The proposed channel assignment algorithm starts by sorting links in the descending order of their link costs. Then, channels are assigned to the links in that order. The proposed algorithm avoids self-interference by not assigning a channel to any link whose incident links have already been assigned channels. In other words, a link is eligible for activation only if it has no active neighbor links. In order to alleviate the effects of cochannel interference, the channel that is assigned to a link is selected based on the sum of link gains between all the interfering senders using the same channel and the receiver of the link. This sum is calculated for each of the channels and the channel with the least associated value is selected for the link.

Resource Allocation:

The main objective of the module is to allocate resources to the different connections such that the minimum rate requirements of each connection are met. The proposed approach requires both the transport (in terms of end-to-end rate allocation) and the physical layer (in terms of channel and power schedule) to be aligned. Coordination between the two layers can be implemented on different timescales: end-to-end rate allocation (through TCP/AQM) on the fast time-scale and incremental channel and power updates on the slow time-scale. Most of the common TCP/AQM variants can be interpreted as distributed methods for solving the optimization network flow problem (determines the end-to-end rates under fixed link capacity). Based on an initial schedule (a simple TDMA link schedule for the first L slots), we run the TCP/AQM scheme until convergence (this may require the schedule to be applied repeatedly). After rate convergence, each node reports the link prices associated with its incoming and outgoing links to gateway where the proposed resource allocation scheme is adopted. On receiving the link prices from the entire set of node, the gateway finds the channels and transmits powers by applying the resource allocation scheme proposed; it then augments the schedule. The procedure is then repeated with this revised schedule.

Admission control module

An admission control strategy is essential to provide protection to the sources that are currently being serviced. In other words, the QoS of existing flows in terms of a minimum rate (being currently provided) cannot be compromised in order to accommodate new incoming flows. Our resource allocation framework can be easily adapted to support admission control.



  • Processor             –        Pentium –IV
  • Speed –     1 Ghz
  • RAM –     256 MB
  • Hard Disk –      20 GB
  • Key Board –     Standard Windows Keyboard
  • Mouse – Two or Three Button Mouse
  • Monitor – SVGA


  • Operating System : Windows XP
  • Programming Language : NS2
  • Tool : CYGWIN


Tae-Suk Kim, Yong Yang, Jennifer C. Hou,Fellow, IEEE,and Srikanth V. Krishnamurthy,Fellow, IEEE “Resource Allocation for QoS Support in Wireless Mesh Networks” – IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, 2013

About the Author