ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

Design of an Efficient QoS Architecture (DEQA) for Mobile Ad hoc Networks

1

R.Mynuddin Sulthani1 and D.Sreenivasa Rao 2, Professor in CSE Dept, Madanapalle Institute of Technology & Science (MITS), Madanapalle, Chittoor (Dt), A.P., India. 2 Professors in ECE Dept, JNTU College of Engg, JNT University, Hyderabad, A.P., India. [email protected], [email protected] power management, security and quality of service (QoS) issues. As the nodes communicate over wireless links, each node should fight against the highly erratic nature of wireless channels as well as interference from other transmitting nodes. These factors make it a challenging problem to maximize data throughput while meeting user-required QoS in wireless ad-hoc networks Owing to mobility of the nodes and interference among nodes, great difficulty in implementing ad-hoc networks comes from frequent route changes. The high packet loss rates and frequent topological changes make the transport layer unstable and limit the amount of traffic that the network can carry. Three well-known problems in ad-hoc networks are the lack of reliable packet delivery due to interference and movement of nodes, limited bandwidth due to channel restrictions, and limited node lifetime due to small battery size [14]. Towards enlarging the life span of the network fairness of access to the network, research in QoS issues in wireless ad-hoc network is generally oriented. Along with it, the issues of connections among heterogeneous nodes are treated. For traffic flows in mobile ad-hoc networks, there is much need of research in offering reliable transport services as well as giving QoS guarantees. Since the size of the ad-hoc networks is huge, good scalability of the QoS architecture is necessary. In the case of preceding works on QoS in ad-hoc networks, the scalability issues are very hardly ever measured. On Call Admission Control (CAC) [2], there are numerous research works in wireless networks. Established end-end probing-based admission control possibly will not complete the delay necessity, because the network size is large. On the other hand, network overload can only be thwarted by CAC based on existing bandwidth not including mobility. In order to permit a flow under mobility the topology may be modified. This is ensuing in modifications of traffic distribution in the network. Therefore, even if

Abstract
With the increasing widespread use of wireless technologies, Quality of service (QoS) provisioning in ad hoc networks remains a challenging task. Good scalability of the QoS architecture is necessary, as the size of the ad hoc networks is huge. In this paper we propose a design of an efficient QoS architecture (DEQA) for mobile ad-hoc networks, which consists of a multi-path routing protocol, a call admission control scheme (CAC) and a congestion control mechanism. The multi-path routing protocol utilizes erasure code techniques for producing replicated fragments for each packet, to enhance reliability. Important fragments can be sent through the paths with strong connectivity and high delivery probability. While it receives an assured number of fragments, destination can recover the original packet. In the CAC scheme, requests are admitted based on the bandwidth availability of the path. To avoid performance degradation due to mobility-triggered congestion, a congestion control mechanism has been developed. Once congestion occurs, the best effort traffic is rate controlled, to free bandwidth for the real-time flows. All these schemes together make the proposed QoS architecture scalable to large network size and mobility. By simulation results, it has been shown that, this architecture allows us to provide reliable QoS in ad-hoc networks increases the available bandwidth, performs load balancing in the network and increases lifetime.

Keywords: Scalable, Congestion, Fragments, Bandwidth, MultiPath 1. Introduction

Reliable,

The ad-hoc wireless network is recognized as an extraordinary case of wireless network, which has no fixed backbone infrastructure. Owing to the above mentioned property, wireless ad-hoc networks can be flexible and rapidly deployed, but also poses significant technical challenges. Numerous challenges include effective routing, medium access,

49

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

there is sufficient bandwidth at the time of flow admission, overcrowding might still take place under mobility. This paper proposes a design of a scalable and reliable QoS architecture for mobile ad-hoc networks. It consists of i) a multi-path routing protocol, ii) a call admission control (CAC) scheme and iii) a congestion control mechanism. Towards enhancing reliability, a multi-path routing protocol, utilizing erasure code techniques for producing replicated fragments for each packet, has been proposed. After establishing the multiple paths, they can be arranged based on their strong connectivity and delivery probability. Important fragments can be sent through the stronger paths among the replicated fragments. The subsequent stronger path can be chosen from the list, if there are some unpredicted faults that happen in the path. While it receives an assured number of fragments, destination can recover the original packet. In the proposed CAC scheme, the calls based on the bandwidth availability of the path are admitted. The QoS routing to propagate bandwidth information throughout the network is adopted. An additional congestion control scheme has been developed to thwart performance degradation as a result of mobility-triggered congestion. To assign free bandwidth for the real-time flows, best effort traffic must be rate controlled once congestion occurs. All these schemes together make the proposed QoS architecture scalable to large network size and mobility.

2. Related Work
In [1], the authors have introduced and evaluated the DLite algorithm, a novel approach to service differentiation in ad-hoc networks, which applies a fair queuing scheme with separate queues for each service class. Late packets of delay constrained classes are dropped in intermediate routers. DLite is easy to implement and requires low computational overhead. It allows for adaptive multimedia applications and permits gradual deployment. In [2], the authors have proposed a scalable QoS architecture for ad-hoc networks. This scheme draws upon the positive aspects of both IntServ and DiffServ, and extends upon the scalable LANMAR routing protocol to support QoS and also it is capable of incorporating mobile backbone networks (MBNs) to further improve the scalability. In [3], the authors have proposed a new QoS framework for MANETsAdaptive Reservation and Pre-allocation Protocol (ASAP). By using two signaling messages, ASAP provides fast and efficient QoS support while maintaining adaptation flexibility and minimizing wasted reservations In [4], the authors have performed a study on the

various queuing schemes for multi-hop wireless networks and examine the fairness and throughput performance of each scheme. Each scheme offers a different degree of fairness. While relatively simple queuing schemes require less hardware and processing budget, they inevitably lack good fairness and performance. In contrast, the scheme that provides fairness requires per-flow (i.e., networklayer flow) queuing. In [5], the authors have presented an agent based scheme for efficient management of radio resources in hybrid wireless networks. Performance of the agent based scheme is measured in terms of successful handover rate between different wireless network architectures (e.g., WLAN, Cellular), and also by the allocated bandwidth to admitted calls. In [6], the authors have discuss a new packet scheduling models for an multi-hop wireless network, and which ensures fair allocation of basic channel service while seeking to maximize spatial reuse. The objective of the above model is to devise effective scheduling disciplines to provide packet-level QoS in terms of throughput, delay and fairness. The authors demonstrate a packetized algorithm that realizes the scheduling model with analytically provable performance bounds. In addition a backoff-based distributed implementation is designed in the above mentioned paper, which closely emulates the ideal centralized algorithm. The above mentioned paper also demonstrates the effectiveness of the devised algorithm through both simulations and analysis. In [7], the authors have proposed a QoS-aware routing protocol that incorporates an admission control scheme and a feedback scheme to meet the QoS requirements of real-time applications. The novel part of the above mentioned QoS-aware routing protocol is the use of the approximate bandwidth estimation to react to network traffic. The above mentioned approach implements these schemes by using two bandwidth estimation methods to find the residual bandwidth available at each node to support new streams. In [8], the authors have performed a study on interference-aware topology control and QoS routing in IEEE 802.11-based multi-channel wireless mesh networks with dynamic traffic. Channel assignment and routing are two basic issues in such networks. Different channel assignments can lead to different network topologies. A novel definition of co-channel interference has been presented in the above mentioned paper. Based on this concept, an effective heuristic for the minimum Interference Survivable Topology Control (INSTC) problem is formally defined and presented which seeks a channel assignment for the given network such that the induced network topology is interference-minimum among all K-connected topologies, which improves

50

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

the system performance by 57% on average in terms of connection blocking ratio. In [9], the authors have discussed an approach to support delay-sensitive multimedia applications over hybrid wireless/wired networks. By the appropriate selection of the AIMD protocol parameters, wireless resources can be efficiently utilized, flow throughput can be maximized under the constraint of the delay outage probability. Simulation results have validated the analysis, demonstrated the feasibility of the approach, and shown that the AIMD protocol can outperform the non-responsive UDP protocol when they are used to support multimedia applications over hybrid networks. In [10], the authors have presented a users' satisfaction factor (USF) defined to quantify quality of service (QoS) for different types of services such as voice, data, and multimedia, as well as for different delay constraints. This USF not only predicts the final delivered QoS during transmission, but also take advantages of the fact that different packets can be decoded at different time in the receivers. Based on this USF, four types of scheduling schemes considering tradeoffs between system performance and individual fairness are presented. The schemes presented explore the time, channel, and multi-user diversity to guarantee quality of service and enhance the network performance.

All these schemes together make the proposed QoS architecture scalable to large network size and mobility.

4. Multipath Reliable Protocol

A. Packet Dispersion Model
To reduce the packet loss, we propose a new packet dispersion mechanism which splits the data packets at the source, into fragments and distributes the fragments on multiple parallel paths. At the destination, packets are reassembled. In order to make this mechanism efficient, we need to use an erasure code technique based on robin dispersal algorithm [13]. The source node breaks up the packet into N fragments of size s, generates K fragments of parity and transmits the total of N+K packets to the destination. For the transmission is to be successful, the destination must receive at least N fragments in at most Tm time units. Important fragments can be sent through the stronger paths among the replicated fragments. The subsequent stronger path can be chosen from the list, if there are some unpredicted faults that happen in the path. The next section describes the process of determining stronger paths, based on connectivity and delivery probability.

B. Nodal Delivery Probability

The data transmission is based on the parameter Node Delivery Index ( NDI ) , which represents the probability that a node correctly delivers data to the destination. Let NDI i denotes the delivery index of a node N i . The value of NDI i initially set to zero and updated whenever there is a message transmission or timer expiration. Whenever a node N i transmits a data packet to another node N K , NDI i should be updated such that

3. Overview of the architecture

In this paper, we propose a design a scalable and reliable QoS architecture for mobile adhoc networks. It consists of i) a multi-path routing protocol, ii) a call admission control (CAC) scheme and iii) a congestion control mechanism. A multi-path routing protocol, utilizing erasure code techniques for producing replicated fragments for each packet, has been proposed to enhance reliability. After establishing the multiple paths, they can be arranged based on their strong connectivity and delivery probability. Important fragments can be sent through the stronger paths among the replicated fragments. The subsequent stronger path can be chosen from the list, if there are some unpredicted fault happens in the path. While it receives an assured number of fragments, destination can recover the original packet. In our CAC scheme, we admit the calls based on the bandwidth availability of the path. We adopt the QoS routing to propagate bandwidth information throughout the network. An additional congestion control scheme has been developed to thwart performance degradation as a result of mobility-triggered congestion. To assign free bandwidth for the real-time flows, best effort traffic must be rate controlled once congestion occurs.

NDI i = (1 ) NDI i + NDI k ,

(1)

Where NDI k is the NDI of node N k , 0 < < 1 , is a constant employed to keep partial memory of historic status. If N k is the destination, then NDI k = 1 , because the message is already delivered to the destination successfully. Otherwise, NDI k < 1 . Clearly, NDI i is always between 0 and 1. Each node maintains a timer t1.If there is no message transmission within an interval of , then the timer t1 expires. The timer expiration indicates that the node couldnt transmit any data during . So NDI i should be updated as

NDI i = (1 ) NDI i

(2)

So from (1) and (2), we arrive that, the node delivery

51

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

index ( NDI ) of node N i is updated as

= (1 ) NDI i + NDI , If there is a

data Transmission = (1 ) NDI i , If there is a Timeout

C. Reliable Data Forwarding

Our Reliable Multipath Protocol (RMP) uses an Multi Path Set (MPS) comprising node-disjoint paths, determined using the AOMDV protocol [11]. The construction of An MPS of node-disjoint paths is made through succeeding calculation of the nodedisjoint, shortest in number of hops, paths, using the route discovery provided network connectivity information. While fresh connectivity information is obtained, RMP efforts to determine new paths usually either proactively or reactively, following the invocation of a route discovery. After the determination of MPS, the source S disperses each outgoing message, adding limited redundancy to the data and dividing the resultant information into pieces, which are transmitted across the MPS routes one piece per route. Even if some of the message pieces are lost or corrupted, successful reception of M out of N pieces allows the reconstruction of the message at the destination. The ratio T = N / M is termed the redundancy factor, and we denote a dispersed message with redundancy T as an ( M , N ) -message. The K packets are subdivided into n non-overlapping sets with M i packets in set i , where M i packets are transmitted on path RI. We will denote with

paths. The advantages of using of dispersion algorithm [12] and erasure code [13] is, simplified message manipulation and reliable data transmission. At the same time, however, the optimization of erasure coding parameters is usually inaccurate because they are calculated according to the current NDI of the source node. In addition, propagating many small messages in the network may incur further processing overhead and inefficiency of bandwidth utilization. In order to avoid the above problems we describe the process of message faulttolerance based data forwarding in the subsequent section.

D. Determining Message Delivery Probability

The Message Delivery Probability (MDP) is used to represent the amount of redundancy and to indicate the importance of a given message. The MDP is defined to be the probability that at least one copy of the message is delivered to the destination by other nodes in the network. We assume that each message copy carries a field to denote the MDP. When a message is generated, its MDP is initialized to be zero. Let us consider a node N i , which is transmitting k copies of a data message j. Let MDP i denote the MDP of message j in the queue of node Ni. The message transmitted to a node Nr is associated with a MDP of
MDP j Nr = 1 (1 MDP j i ) (1 - NDIi )  (1 MDP j m )
m =1 k

and the MDP of the message at node Ni is updated as

MDP j i = 1 (1 MDP j i )  (1 - MDP j m )
m =1 k

m = {M 1 , M 2 ,..., M n } the allocation vector for

T

the connection over the n paths in the network. The probability that the packet can be reconstructed is the probability that more than M of the fragment transmissions succeeded, or equivalently that less than K transmissions failed. We can express network reliability or probability of success as:

Psucc = Pr[ Number of Re ceived Fragments M ]

The efficiency of path diversification is defined as:

The above process repeats at each time when message j is transmitted to another node. In general, the more times a message has been forwarded, the more copies of the message are created, thus increasing its delivery index. As a result, it is associated with a larger MDP. So the messages with larger MDP are dispersed and sent through the best paths with high ANDI.

Effective Throughput M = Actual Throughput M +K

5. Call Admission Control (CAC)

Apart from the multipath reliable routing protocol, our QoS architecture provides Call admission Control (CAC). When a new request with certain bandwidth requirement comes, the source will perform admission control following the steps described below [2]. The source node first consults the local routing table. If there is enough bandwidth to the destination, the source node further checks required bandwidth. If it is less than the minimum bandwidth, the flow can be admitted.

We ignore the overhead introduced by the rest of the network. While transmitting dispersed data, the status of the MPS routes are continuously updated based on each nodes delivery index. For this, for each path, an average node delivery index (ANDI) is calculated based on the NDI of each node in the path. For each successful or failed data, the value of ANDI of the corresponding route is increased or decreased. In case the existing route encounters some unexpected link or route failure, the algorithm selects the path with the next highest ANDI, from the list of selected disjoint

52

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

If it is more than the maximum bandwidth, the flow is rejected. If it is between the minimum and maximum bandwidth, a probing packet is sent to the destination node to obtain the exact available bandwidth at the destination. Based on the results, it there is available bandwidth, the flow is accepted; otherwise it is rejected.

6. Congestion Control
In mobile ad-hoc networks, call admission control at source nodes alone cannot guarantee QoS since the topology may change after flows are admitted. Network congestion can still occur frequently under mobility. Thus, congestion control is needed to provide QoS in such situations. When congestion occurs, we would like best-effort traffic to first reduce their transmissions rate to give bandwidth to real-time flows.

In our simulation, 50 mobile nodes move in a 1000 meter x 1000 meter rectangular region for 50 seconds simulation time. We assume each node moves independently with the same average speed. All nodes have the same transmission range of 250 meters. In this mobility model, a node randomly selects a destination from the physical terrain. It moves in the direction of the destination in a speed uniformly chosen between the minimal speed and maximal speed. After it reaches its destination, the node stays there for a pause time and then moves again. In our simulation, the minimal speed is 5 m/s and maximal speed is 10 m/s. The simulated traffic is Constant Bit Rate (CBR). Our simulation settings and parameters are summarized in table 1 No. of Nodes Area Size Mac Radio Range Simulation Time Traffic Source Packet Size Speed Flows Rate 50 1000m X 1000m IEEE 802.11 250m 50 sec CBR 512 10m/s 2,4,6,8 and 10 250,500,750, & 1000 Kb

A. Network Congestion Detection

For congestion control, we need to first detect the congestion in the network. In multi-hop wireless networks, accurate congestion detection of a neighborhood is difficult. In this paper, congestion detection in a nodes neighborhood is performed by monitoring the wireless channel utilization ratio, which can be obtained by the adaptive bandwidth estimation scheme. When the channel utilization ratio is larger than a pre-defined threshold value, it can be assumed that this nodes neighborhood is entering a congested state.

Table 1: Simulation Parameters

B. Performance Metrics
The proposed DEQA is compared with the SMT [13] scheme. The performance is evaluated mainly, according to the following metrics. i. ii. Average end-to-end delay: The end-to-enddelay is averaged over all surviving data packets from the sources to the destinations. Average Packet Delivery Fraction: It is the ratio of the number .of packets received successfully and the total number of packets transmitted. Aggregated Throughput: We measure aggregated throughput in terms of no. packets received. Fairness: For each CBR flow, we measure the fairness as the ratio of throughput of each flow and total no. of flows. Packet Loss: We measure the packet loss , which is the no. of packets lost per unit time. Blocking Probability: We measure the blocking probability as the ratio of rejected requests per total no. of requests. Overhead: It is the overhead occurred due to control packets exchange.

B. Rate Control of Best Effort Traffic

By applying rate control to best-effort traffic, congestion control can be performed. The best-effort traffic may use unconsumed bandwidth of the realtime flows. But whenever a new real-time flow arrives, the best effort traffic flows are forced to free the bandwidth for the new real-time flow. The starvation of best effort flows can be avoided by reserving a small fraction of the bandwidth to besteffort traffic at all times. The basic idea of the scheme is, when there is no congestion, the best-effort traffic increases its rate slowly. Once congestion is detected, the ECN bits of best-effort traffic will be marked and transmitted to the source nodes. On receiving this ECN packet, source nodes will reduce their sending rate.

iii.

iv.

7. Experimental Results
A. Simulation Model and Parameters
We use Network Simulator (NS2) to simulate our proposed algorithm. In our simulation, the channel capacity of mobile hosts is set to the same value: 2 Mbps. We use the distributed coordination function (DCF) of IEEE 802.11 for wireless LANs as the MAC layer protocol. It has the functionality to notify the network layer about link breakage.

v. vi.

vii.

53

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

C. Results
In this section simulation results are presented. 1. Rate In the first experiment, the transmission rate is varied as 250kb; 500kb1000Kb and the above metrics are measured. Fig1 shows the result of throughput in terms of packets. From the figure, it can be seen that DEQA scheme outperforms SMT, by gaining more packets. Fig 2 shows the PDF for DEQA and SMT. Clearly DEQAs PDF is significantly more than that of SMT. Fig 3 gives the end-to-end delay values. Due to its routing policy, DEQA has less delay when compared to SMT. Fig. 4 presents the packets dropped for both the schemes. Because of its reliability, DEQA has less packet drops than SMT. The blocking probability results for various bandwidth requests are presented in figure 5. Because of its call admission control policy, DEQA has reduced blocking probability, when compared with SMT. The fairness of all the flows are given in figure 6. Since the DEQA scheme allocates fair bandwidth for its flows, it has increased fairness when compared to SMT. Fig 7 shows that the Overhead required for SRQA is less compared overhead of SMT.
T hro ug hp ut 70 00 60 00 50 00 40 00 30 00 20 00 10 00 0 2 50 5 00 75 0 10 00 R ate(kb )

more packets. Fig 9 shows the PDF for DEQA and SMT. Clearly DEQAs PDF is significantly more than that of SMT. Fig 10 gives the end-to-end delay values. Due to its CAC and congestion control schemes, DEQA has less delay when compared to SMT. Fig. 11 presents the packets dropped for both the schemes. Because of its reliability, DEQA has less packet drops than SMT. The blocking probability results for various bandwidth requests are presented in figure 12. Because of its call admission control policy, DEQA has reduced blocking probability, when compared with SMT. The fairness of all the flows are given in figure 13. Since the DEQA scheme allocates fair bandwidth for its flows, it has increased fairness when compared to SMT. Fig 14 shows that the Overhead required for SRQA is less compared overhead of SMT.
End-to-End Delay 3 2.5 Delay(s) 2 1.5 1 0.5 0 250 500 750 1000 Rate(kb) DEQA SMT

Fig.3 Rate Vs Delay

DEQA SMT
50000 40000 Drop 30000 20000 10000 0 250
P a c k e tD e liv e r y F r a c tio n

Packets

R a te V s D r o p

DEQA SMT

Fig.1 Rate Vs Throughput

500

750

1000

R a te

120 100 PDF(%) 80 60 40 20 0 250 500 750 1000 R a te (k b ) DEQA SMT

Fig.4 Rate Vs Packets Dropped

Rate Vs Blocking Probability Blocking Probability 0.5 0.4 0.3 0.2 0.1 0 250 500 Rate 750 1000 DEQA SMT

Fig.2 Rate Vs PDF

2. Flows In the second experiment, the no. of data flows is varied as 2, 4, 6,8,10 and the same metrics are measured. Fig8 shows the result of throughput in terms of packets. From the figure it can be seen that, DEQA scheme outperforms SMT, by gaining

Fig.5 Rate Vs Blocking Probability

54

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

Rate Vs Fairness 1.2 1 0.8 0.6 0.4 0.2 0 250 500 Rate 750 1000
10000 8000 Drop 6000 4000 2000 0 2

F lo w V s D r o p

Fairness

DEQA SMT

DEQA SMT

6 F lo w

10

Fig.6 Rate Vs Fairness

Rate Vs Overhead
14000 12000 10000 8000 6000 4000 2000 0 250 500 Rate 750 1000

Fig.11 No. of Flows Vs Packets Dropped

F lo w V s B lo c k in g P r o b a b ility Blocking Probability 0 .3 0 .2 5 0 .2 0 .1 5 0 .1 0 .0 5 0 2 4 6 F lo w 8 10

DEQA SMT

Overhead

DEQA SMT

Fig.12 No. of Flows Vs Blocking Probability

Flow V s Fairness

Fig.7 Rate Vs Overhead

Throughput 7000 6000 5000 4000 3000 2000 1000 0 2 4 6 Flows 8 10

0.5 Fai 0.4 rne 0.3 ss 0.2 0.1 0 2 4 6 Flow 8 10

DEQA SMT

packets

DEQA SMT

Fig.13 No. of Flows Vs Fairness

F lo w V s O v e rh e a d 10000 8000 6000 4000 2000 0 2 4 6 F lo w 8 10 DEQA SMT

Fig.8 No. of Flows Vs Throughput

PacketDeliveryFraction 120 100 80 60 40 20 0 2 4 6 Flows 8 10

PDF(%)

DEQA SMT

Overhead

Fig.14 No. of Flows Vs Overhead

9. Conclusion
In this paper, a joint design of reliable QoS architecture for mobile adhoc networks has been designed which consist of a multipath routing protocol, a call admission control scheme and a congestion control mechanism. In the reliable multipath routing protocol, dispersion and erasure code techniques are utilized for producing replicated fragments for each packet, to enhance reliability. Then messages with good delivery probability are identified and transmitted through the paths with high average node delivery index. While it receives an assured number of fragments, destination can recover the original packet. Next, a call admission control (CAC) scheme has been developed, in which, the

Fig.9 No. of Flows Vs PDF

E nd-to-E nd D elay 1 0.8 delay(s) 0.6 0.4 0.2 0 2 4 6 Flow s 8 10 DEQA SMT

Fig.10 No. of Flows Vs Delay

55

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

calls are admitted based on the bandwidth availability of the path. Finally, a congestion control mechanism has been developed to avoid performance degradation due to mobility-triggered congestion. Once congestion occurs, the best effort traffic is rate controlled, to free bandwidth for the real-time flows. By simulation results, it has been shown that, this combined QoS architecture achieves good throughput, reduces packet loss, increases available bandwidth, performs load balancing in the network and increases lifetime.

References
[1] Michael Gerharz, Christian de Waal, Matthias Frank, "A Practical View on Quality-of-Service Support in Wireless Ad Hoc Networks". In Proc. of the 3rd IEEE Workshop on Applications and Services in Wireless Communications and Mobile Computing 2002. [2] K. Xu, K. Tang, R. Bagrodia, M. Gerla, and M. Bereschinsky, Adaptive Bandwidth Management and QoS Provisioning in Large Scale Ad Hoc Networks, Proc. Military Comm. Conf. (MILCOM '03), Oct. 2003 [3] Xue, J. Stuedi, P. Alonso, G.,"ASAP: an adaptive QoS protocol for mobile ad hoc networks",14th IEEE proceedings on Personal, Indoor and Mobile Radio Communications, 2003. PIMRC 2003. [4] Jangeun Jun Sichitiu, M.L.,"Fairness and QoS in multihop wireless networks", IEEE 58th Vehicular Technology Conference, 2003. VTC 2003-Fall. [5] R. Babbar, A.O. Fapojuwo, B. Far, Agent-based resource. Management in hybrid wireless networks, accepted for presentation, IEEE Canadian Conference on Electrical and Computer Engineering, 2004 [6] Haiyun Luo, Songwu Lu, Vaduvur Bharghavany, Jerry Cheng, Gary Zhongy, "A Packet Scheduling Approach to QoS Support in Multihop Wireless Networks", ACM Journal of Mobile Networks and Applications (MONET) Vol. 9, No. 3, June 2004 [7] Lei Chen,Wendi B. Heinzelman, "QoS-Aware Routing Based on Bandwidth Estimation for Mobile Ad Hoc Networks", IEEE Journal on selected Areas in Communications, vol. 23, no. 3, March 2005. [8] J. Tang, G. Xue, and W. Zhang, "Interferenceaware topology control and QoS routing in multichannel wireless mesh networks,", Proceedings of the 6th ACM international symposium on Mobile ad hoc networking and computing, May 2527, 2005. [9] Lin Cai, Xuemin (Sherman) Shen, Jon W. Mark, and Jianping Pan, "QoS Support in Wireless/Wired Networks Using the TCPFriendly AIMD Protocol", IEEE Transactions On

Wireless Communications, VOL. 5, NO. 2, FEBRUARY 2006. [10] Zhu Han; Xin Liu; Z. Jane Wang; K. J. R. Liu, "Delay sensitive scheduling schemes for heterogeneous QoS over wireless networks", IEEE Transactions on Wireless Communications, Volume 6, Issue 2, Feb. 2007 Page(s):423 428 [11] Mahesh K.Marina, Samir R.Das Ad hoc Ondemand Multipath Distance Vector Routing, ACM Sigmobile Mobile Computing and Communications Review Volume 6 , Issue 3, July 2002. [12] Panagiotis Papadimitratos, Zygmunt J. Haas, Secure Data Communication in Mobile Ad Hoc Networks, IEEE Journal On Selected Areas In Communications, Vol. 24, No. 2, February 2006. [13] Panagiotis Papadimitratos, Zygmunt J. Haas Secure message transmission in mobile ad hoc networks , 2003 Elsevier, Ad Hoc Networks 1 (2003) 193209. [14] P. Djukic, "Optimum Resource Allocation in Multipath Ad Hoc Networks", MASc Thesis, University of Toronto, 2003

56

ICGST-CNIR Journal, Volume 8, Issue 2, January 2009

R.Mynuddin Sulthani received the B.E., degree in Electronics and Communication Engineering from Osmania University, Hyderabad, India in 1992, and M.Tech., in Digital Systems & Computer Electronics from JNT University, Hyderabad, India. He is pursuing his PhD work on Quality of Service (QoS) in wireless Networks, under JNT University, Hyderabad, India. His research interests are in the areas of wireless/mobile networks with emphasis on the QoS guarantee and adaptation, connection and mobility management, and wireless MAC protocols. He is currently working as a Professor & HOD in CSE Department, Madanapalle Institute of Technology & Science (MITS), Madanapalle, which is affiliated to JNTU, Hyderabad, India. He has 7 years of Industrial experience and 10 years of teaching experience. He has attended an international conference ICCCN-2008 and several Workshops conducted by various organizations in India. He has organized national level seminars, workshops and refresher courses in MITS.

D.Sreenivasa Rao received the B.Tech., degree in Electronics and Communication Engineering in 1986 from Nagarjuna University, India and M.E in Digital Systems in 1994 from Osmania University, Hyderabad, India. He was awarded the PhD degree in Computer Networks in 2004 from University of Hyderabad, India. He is a Professor in ECE Department, JNTU College Of Engineering, JNT University, Hyderabad, India. He has 20 years of teaching experience. His research interest is in the area of communications and computer networks Presently 12 research students are working under his guidance. He has 22 publications in various National, International Conferences and Journals. He has 20 years of teaching experience. He has attended more than 10 Short Term Courses, Summer Schools, and Workshops, conducted by various organizations. He has organized workshops and refresher courses. He has chaired sessions at various national conferences. He is advisory committee member for GNIT, Hyderabad. He is also governing body member for Syed Hashim College of Science & Technology, member of JNTU forum for Science & Society and Coordinator for campus networking at JNTU CE, Hyderabad

57