Art

A generic mobile node architecture for multi-interface heterogenous wireless link layer

Description
A generic mobile node architecture for multi-interface heterogenous wireless link layer
Categories
Published
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  A Generic Mobile Node Architecture forMulti-Interface Heterogenous Wireless Link Layer Rafaa TAHAR HANA Research GroupUniversity of Manouba, TunisiaEmail: rafaa.tahar@hanalab.org Abdelfettah BELGHITH HANA Research GroupUniversity of Manouba, TunisiaEmail: abdelfattah.belghith@ensi.rnu.tn Rafik BRAHAM Prince Research UnitUniversity of Sousse, TunisiaEmail: rafik.braham@ensi.rnu.tn  Abstract —In this paper, we first propose a new and openarchitecture for mobile wireless nodes using multiple interfaces.This modular and flexible architecture aims to offer to industrialand research communities several possibilities to investigate andsimulate multiple wireless protocol aspects, algorithms and issues.To validate our architecture and to show its advantages, wechoose to implement a multiple interface WLAN protocol inthe widely used OMNeT++ Discrete Event Simulator [1] basedon the MMAC Protocol [2] and its derivatives [3]. While usingmultiple interfaces leads certainly to an enhancement of theentire network performance, the question remains to specifyalgorithms and mechanisms that realize a compromise betweenpower consumption and performance efficiency (Throughput,End-to-End Delay).Secondly, we present a modified version of the PSM-MMACProtocol [3] that we called the Adaptive ATIM Multi-InterfaceMAC Protocol (AA-MIMP). Using our proposed open architec-ture, we show that AA-MIMP realizes a substantial enhancementof QoS parameters and performances, yet demonstrating asubstantial improvement of power consumption. I. I NTRODUCTION In the last few years the deployment of wireless tech-nologies, especially WLANs based on the IEEE-802.11 stan-dard [4], [5], has grown exponentially. However, this tech-nology still suffers from certain inefficiencies such as lowthroughput, inefficient power conservation, lack of real timeconsideration for peer-to-peer and interactive applications in-cluding VoIP, VoD, Streaming, etc. Several recent researchworks tried to propose enhancements and thrived to offer somesolutions. Of particular interest to our present work are thosebased on the efficient use of the available frequency spectruminstead of just being limited to a unique radio channel ascurrently deployed. The major contribution of this work isto propose a new architecture of a mobile node that permitsthe design and implementation of algorithms and protocolsallowing the enhancement of wireless communication. Wealso demonstrate how the new architecture could implementa multi-interface protocol for IEEE 802.11 WLANs. Thispaper is organized as follow. In the first section we presentdetails and key features of the new architecture. In the secondsection we show how to implement a modified version of the PSM-MMAC Protocol [3] that we call Adaptive ATIMMulti-Interface MAC Protocol (AA-MIMP). In the last sectionwe will conduct a performance evaluation to show the en-hancement of QoS parameters (throughput and end-to-end datadelay) and a substantial improvement of power consumption.II. M OTIVATIONS AND RELATED WORK Many works dealt with the multi-channel / multi-interfacewireless communications [6], [7], [8]. Some of them proposedto set up ”virtualization” techniques at the Link layer (MACand LLC) [9]. These techniques aim to maintain specific datastructures in order to realize suitable mapping between thelogical information of the Network Layer and the physicalinformation of the Link and Physical layers. Some works dealwith multi-channel / multi-interface heterogeneous wirelesscommunications issues, but to the best of our knowledge,only the Hyacinth [10] architecture tried to propose a multichannel hybrid technology IEEE 802.11/IEEE 802.16 [11]integration. The Hyacinth architecture proposes to use multipleIEEE 802.11 channels to define ”aggregate coverage areas” formobile nodes and uses IEEE 802.11 or IEEE 802.16 technolo-gies to ensure the connectivity of these areas to wired network.But this architecture did not deal with nodes that combinemultiple technologies which is the case nowadays with mobilenodes having multiple built-in wireless technologies: IEEE802.11, IEEE 802.11p [12] and IEEE 1609.x (VANET), IEEE802.15 [13], ...In this paper, we will focus solely on the multi-interface ap-proach for several reasons. The NIC (Network Interface Card)has a low cost and still decreasing. This approach offers moreflexibility in implementation and a relative relaxation fromthe constrained aspects of synchronization between clocks.Research and implementations already conducted at HANAResearch Group showed the advantages of the Multi-interfaceapproach [14], [15], [16].III. P ROPOSED  M ULTI  I NTERFACE  N ODE  A RCHITECTUREFOR  W IRELESS  N ETWORKS In order to take advantage of the available frequency spec-trum within a wireless network, and regarding the low costof wireless NIC, a simple idea consists of equipping eachmobile node with multiple NICs. But this NIC multiplicitywill induce many complex issues. The most important of themare : NIC management (especially for heterogeneous NICs),use of multiple wireless technologies, overhead optimization,power conservation, clocks drift, interface selection and traffic 978-1-4673-5265-9/12/$31.00 c  2012 IEEE  distribution.To solve these problems, we propose a new mobile node ar-chitecture that must also guarantee that protocol modificationsshould be minimal and flexible to authorize future works forboth industrial and research communities. The key features of our model could be resumed in these points: •  The implementation of a Layer-2 virtualization tech-nique called  VML  (Virtual MAC Layer) by deriva-tion/overloading of the LLC sub-layer. Indeed this VMLwill ensure an abstraction of the NIC multiplicity regard-ing upper layers (IP and upper layers); •  An interface is represented by a  PM-Bloc  composed of a Physical and MAC layer which makes our architecturephysically aware (no modification of the Physical Layer); •  We implement necessary but not exhaustive data struc-tures and algorithms for interface selection policies, NICstatus identification (sender, receiver, inactive); •  Our implementation is integrated and tested within thewidely used OMNeT++ Discrete Event Simulator v4.1.  A. Multi Interface Mobile Node Architecture In mostly used network discrete event simulators (NS-2,J-SIM, OMNeT++, ...), a classic mobile node is composedof : A Physical layer, A MAC sub-layer that implements theprotocol state machine, A Management sub-layer (LLC) thatimplements management procedures, complementary modulesthat implement wireless communications specific aspects suchas : mobility, propagation models, cross-layering features, ...and upper layers modules (IP, Transport, Applications). Whenwe deal with efficient use of different orthogonal wirelesschannels there are mainly two modes : multi-channel andmulti-interface. The difference between these two modes couldbe resumed as follows: •  Multi-interface mode allows the simultaneous use of thetotal frequency spectrum •  Multi-channel mode reduces channel noise and interfer-ence •  Since every interface is tuned to a given channel, multi-interface MAC layer management is more flexible andeasy because it does not need channel switch, scan, ... •  Multi-interface LLC layer management is more complexsince other procedures are necessary such as : multipleMAC address management, power conservation, ...Since multi-interface mode is more efficient, we conducted acomplete study and we propose a new mobile node architec-ture, as shown in Figure 1. It is composed of : •  ”PM-BLOCs” that associate a MAC sub-layer state ma-chine and a physical module •  A VML (Virtual MAC Layer) that substitute the LLCsub-layer.Notice that our architecture could also be generalized to covermulti-channel issues by introducing ”switching time” betweenchannels and activating only one PM-Bloc at a time.Fig. 1: New Mobile Node Architecture 1) Physical and MAC Sub-Layer BLOC (PM-Bloc):  Thehard coupling between the physical layer and the MAC sub-layer is due to two objective constraints. They are : •  MAC sub-layer is a logical representation of the phys-ical underlying layer by means of MAC protocol statemachine; •  There is no more distinction between the modules since inthe multi-interface mode each NIC is tuned to a specificchannel.Recall that since we choose to implement the PSM-MMACprotocol, some modifications are introduced to the control PM-Bloc. Other blocs are left as-is. 2) VML : Virtual MAC Layer:  To handle the multiplicity of NICs (PM-Blocs), it is necessary to derive/overload the LLCsub-layer in order to deal with: •  The abstraction (virtualization) of multiple NICs regard-ing upper layers; •  Multiple NICs management (Management Frames likeATIM, channels identification, PSM, ...); •  Traffic distribution.Our new architecture which combines VML and PM-Blocsto describe a multi-interface mobile node could be used tostudy multiple scenarios such as Wireless Distribution Sys-tems (WDS) [15], [16], multi-interface Ad-hoc WLANs andVANET. As an application of this architecture, we choose toimplement and enhance the PSM-MMAC protocol [3] basedon both the MMAC [2] and the DCA protocol [17]. Sincethis protocol family is based on the idea that all nodes mustat least have 2 NICs (one dedicated to control and the othersare for data), we introduced solely some modifications to thecontrol PM-Bloc. Data PM-blocs are not modified.IV. A DAPTIVE  ATIM M ULTI -I NTERFACE  MAC P ROTOCOL (AA-MIMP)We now propose an improved version of the PSM-MMACin the quest to enhance power saving and improve overall per-formances. We implement the AA-MIMP using the proposednew architecture. The AA-MIMP enhances the PSM-MMACprotocol [3] and is based on both the MMAC [2] and theDCA protocol [17] in order to enhance QoS parameters and  to reduce power consumption. The AA-MIMP is dedicatedto WLAN in which nodes are in Line of Sight (LoS) andcould be heterogeneous (number of NICs varies from a nodeto another). Unlike the PSM-MMAC Protocol that uses acomplex probabilistic model to estimate the size of the ATIMWindow, our protocol is based on the standard DCF functionand uses control channel sensing to decide or not if the ATIM / ATIM-ACK Frames exchange is terminated or not. Datastructures maintained and information exchanged between thecontrol PM-Bloc and VML allow us to determine the exactstate of all the network during the current Beacon Interval in afully distributed manner. Like the DCA Protocol, our protocolis receiver side channel selection oriented but does not add theRES frame to indicate to neighbors which channel is selectedfor the transmission. Instead, it piggybacks this decision inthe ATIM-ACK frame. Our Protocol is based on the standardtiming structure proposed in the PSM and tries to solve thefollowing problems : •  Dynamic ATIM Window Size estimation regarding thesubmitted traffic; •  Interface selection regarding a traffic distribution specificpolicy (random, load balancing, threshold, ...); •  QoS parameters enhancement; •  Power conservation.Unlike the complex probabilistic model for ATIM WindowSize Estimation proposed in the PSM-MMAC Protocol basedon Active Links (AL) and which suppose that traffic loads sub-mitted to the network are constant, our estimation algorithmis most adaptive and is based on the observation of Packet’sQueue Size at each start of a Beacon Interval and we did notmake any assumption on traffic loads which is the case of WLANs, WDS, ... So our algorithm is based on the followingkey features : •  At the start of every Beacon Interval, all control PM-Blocs are ”awake” and still ”awake” for at least the restof the ATIM Window; •  Only nodes that have traffic will send an ATIM Frameusing the DCF function. So in the ATIM Frame body amobile node will send the following information: PacketNumber (actual Queue Size), Mean Packet Size andavailable Data Channel List; •  Receiver node selects the data channel regarding a pre-defined selection policy (common for all nodes duringthe current Beacon Interval, in our case a load balancingpolicy) and transmits the selected channel number in theATIM-ACK Frame from the common subset channelswith the sender. The choice of this receiver orientedselection policy is due to the fact that sender and receivermobile nodes may have heterogeneous capabilities; •  When the sender receives the ATIM-ACK Frame it willimmediately send its traffic on the dictated data channel.If the dictated channel is the control PM-Bloc, the sendermust delay transmission until the end of the ATIMWindow Estimation Algorithm; •  Estimated ATIM Window Size is decided when anamount of slot time equal to the upper bound of thesize of the current contention interval, that we havecalled  backoffFactor , without sensing any activity onthe control channel or when the bound maximal valueof the ATIM Interval, as mentioned in the PSM standardequal to 0.02 sec, is reached; •  At the end of the ATIM Estimated period, all nodes knowexactly the status of every channel and which PM-Blocsare active and those which must go into ”doze” mode.V. P ERFORMANCE  E VALUATION  A. Definitions and notations We now discuss the AA-MIMP QoS parameters, namely thethroughput and the end-to-end delay. We also discuss DutyCycle and Per bit Power Consumption formally defined asfollows : •  Throughput :  The quantity of information per time unitdelivered successfully to all destination stations. That is: Throughput  =  nbPckt  ∗  pcktSizesimDuration  (1)where  nbPckt   denotes the number of packets receivedsuccessfully by all destination stations.  pcktSize  denotesthe size of a packet and  simDuration  denotes the simu-lation duration in seconds. •  End-to-end delay :  The average end-to-end delay of allpackets delivered successfully to all destination stations.That is : Delay  =  DestArriveTime  −  SrcGenerationTime (2)where  DestArriveTime  denotes the time at which thepacket arrives at the destination station and  SrcGener-ationTime  denotes the generation time of this packet atthe source station. •  Duty Cycle :  We define duty cycle as the average amountof unused transmission time periods of active channelsper beacon interval. That is: dutyCycle  =  BeaconTime ()  −  LastPacketTime () (3)where  BeaconTime()  denotes the current beacon time and  LastPacketTime()  denotes the time at which the last DATAor ACK packet is sent or received. •  Per bit Power Consumption :  We define it as the averagequantity of energy consumed by all node channels dividedby the network throughput. We assume that a node hasfour energy states: Idle (1.0 W), transmission (1.8 W),receive (1.3 W) and doze (0.05 W).  B. Simulation scenario Table I presents general parameters for our simulation.  TABLE I: General Parametrs Parameter Value PYH and MAC bitrate 2 MbpsData channel number 1, 2 and 4 channelsBeacon Interval 0.1 secATIM Max Value 0.02 secFlow Number 1, 5, 10, 15 and 20 FlowsPacket Size 512 BytesPacket tic 0.01 secSimulation Time 200 sec (2000 Beacon Periods)Transmission range 300 meters Fig. 2: Estimated ATIM Window Size vs. Flow Number C. ATIM Window Size Estimation Figure 2 illustrates the Estimated ATIM Window size vs.Number of flows and shows that our estimation is independentof the number of data channels. The figure Fig 3 shows thenumber of ATIM / ATIM-ACK frame exchange succeededduring the ATIM Interval. This number decreases while num-ber of flows increases which is due to high control channelcontention.Fig. 3: Correctly received ATIM  D. Throughput Parameter  Figure 4 shows Throughput vs. Number of flows. Bycomparison with the PSM-MMAC protocol we notice thatour protocol gives better performance with 1 and 2 datachannels for 10 flows and 10 packets/beacon interval and sameperformance with 4 data channels. We notice that when thenumber of flows is equal or higher than 15, a degradationis noticed. This could be explained by both the increasingnumber of unsuccessful ATIM / ATIM-ACK frame exchange,as shown by Figure 3, and DATA Collision number thatincreases due to large channel contention.We notice the sameFig. 4: Throughput vs. Flow Numberphenomenon in the PSM-MMAC protocol but for a numberof flows equal or higher than 4.  E. ATIM Window Size Estimation vs. Throughput  To validate our ATIM Window Size estimation regarding thethroughput parameter we measure the Throughput function of ATIM Window Size. So we will consider 10 flows with 1 datachannel (10 packets per beacon interval) and we will vary theATIM Window Size from 0.005 sec to 0.025 sec. As we canFig. 5: Throughput vs. ATIM Window Sizesee in the Figure 5, the maximal throughput is obtained for anATIM Window Size value equal to 0.0135 sec, which is thevalue obtained by our estimation as shown by Figure 2. F. End-to-End Delay parameter  Figure 6 shows the End-To-End DATA Delay vs. Numberof flows. We notice that with 4 interfaces, End-To-End DATAFig. 6: End-to-End Delay vs. Flow NumberDelay is less than 100 ms for 10 flows (exactly 91.67 ms). Thismeans that multimedia communications are possible. Figure 7is a zoom for 1 and 5 flows for more clarity.  Fig. 7: End-to-End Delay vs. Flow Number (Zoom) G. Duty Cycle Parameter  Figure 8 shows the duty cycle vs. the flow number. WeFig. 8: Duty Cycle vs. Flow Numbernotice that when the number of flows increases and the numberof channels is reduced the duty cycle decreases, which meansthat channel’s capacity is consumed. But we notice a wasteof time when the number of data channels increases. This isdue to unsuccessful transmissions and also to the fact that theBeacon Interval is miss-sized for current transmissions. Thiscould be solved by a good estimation of the Beyond ATIMWindow Size.  H. Per bit Power Consumption Figure 9 shows better power consumption performance ob-tained by correct powering on used and off unused interfaces.We notice that for a low network load (number of flows lessFig. 9: Per bit Power consumption vs. Flow Numberthan or equal to 5), the per bit power cost is minimal for 1data channel. For a medium network load (number of flowsbetween 5 and 10), the energy cost of a bit is minimal for2 data channels. For a high network load (number of flowsgreater than or equal to 15), energy cost per bit is minimalfor 4 data channels. This means that our intelligent powerconsumption approach realizes a good compromise betweenQoS parameters and power consumption and outperforms thePSM-MMAC Protocol.VI. C ONCLUSION Adaptive ATIM Multi-Interface MAC Protocol (AA-MIMP)shows that our new architecture of mobile node is valid,flexible and open to other additions and improvements. OurProtocol realizes a good compromise between QoS param-eters (Network Throughput and End-to-End DATA Delay)and power consumption by intelligent turning on used andoff unused interfaces. This result is confirmed by extensivesimulations which show that the AA-MIMP outperforms thePSM-MMAC protocol in terms of QoS parameters and powerconsumption. While the duty cycle parameter investigationshows that AA-MIMP could be enhanced but needs furtherresearch work. More investigations are also underway toextend the current work to multi-hop. We will also deriveour architecture for VANET communication to demonstrateits validity and its generic character as well.R EFERENCES[1] http://www.omnetpp.org/, “The omnet++ community v-4.1,” Tech. Rep.[2] J. So and N. H. Vaidya, “Multi-channel mac for ad hoc networks :Handling multi-channel hidden terminals using a single transceiver,”  InProc. of Mobihoc, Roppongi, Japan , 2004.[3] J. Wang, Y. Fang, and D. Wu, “A power-saving multi-radio multi-channelmac protocol for wireless local area networks,”  In Proc. of Mobihoc, Roppongi, Japan , 2004.[4] IEEE, “Ieee 802.11 : Part 11 : Wireless lan medium access control macand physical layer phy specifications,” Tech. Rep.[5] ——, “Ieee 802.11b : Part 11 : Wireless lan medium access controlmac and physical layer phy specifications : Higher-speed physical layerextension in the 2.4 ghz band,” Tech. Rep.[6] J. So and N. H. Vaidya, “Routing and channel assignment in multi-channel multi-hop wireless networks with single-nic devices,” Tech. Rep.[7] A. Raniwala, K. Gopalan, and T. cker Chiueh, “Centralized channelassignment and routing algorithms for multi-channel wireless mesh net-works,”  ACM Mobile Computing and Communications Review (MC2R) ,vol. 8, no. 2, 2004.[8] J. So, “Design and evaluation of multi-channel multi-hop wirelessnetwork,” Ph.D. dissertation, Urbana University, Illinois, USA, 2006.[9] R. Chandra, P. Bahl, and P. Bahl, “Multinet: Connection multiple ieee802.11 networks using a single wireless card,”  IEEE INFOCOM  , 2004.[10] A. Raniwala and T. cker Chiueh, “Architecture and algorithms for an ieee802.11-based multi-channel wireless mesh network,”  IEEE INFOCOM  ,2005.[11] IEEE, “Ieee 802.16 : Ieee standard for local and metropolitan areanetworks part 16: Air interface for broadband wireless access systems,”Tech. Rep.[12] ——, “Ieee 802.11p-2010 : Ieee standard for information technology–telecommunications and information exchange between systems–localand metropolitan area networks–specific requirements part 11: Wirelesslan medium access control (mac) and physical layer (phy) specificationsamendment 6: Wireless access in vehicular environments,” Tech. Rep.[13] ——, “Ieee 802.15.4-2006 : Ieee standard for information technology–local and metropolitan area networks– specific requirements– part 15.4:Wireless medium access control (mac) and physical layer (phy) speci-fications for low rate wireless personal area networks (wpans),” Tech.Rep.[14] A. Dhraief, N. Montavont, J. Bonnin, and A. Belghith, “Ns-2 based sim-ulation framework to evaluate the performance of wireless distributionsystem,”  10th Communications and Networking Simulation Symposiumlocated with Spring Simulation Multiconference (SpringSim’07) , 2007.
Search
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks