In Press. INSTITUTE of Electrical and Electronics Engineers

Transcription

1 CommIT (Communication & Information Technology) Journal 12(1), 35 42, 2018 Performance Analys On IEEE ah Standard Enhanced Dtributed Channel Access Mechanm Ana Oktaviana 1, Doan Perdana 2, and Ridha Muldina Negara Telecommunication Engineering Department, Faculty Electrical Engineering, Telkom University 1 anaoktaviana@student.telkomuniversity.ac.id, 2 doanperdana@telkomuniversity.ac.id, 3 ridhanegara@telkomuniversity.ac.id Abstract IEEE ah a new task group on IEEE standard designed to work on 900 MHz. It a range communication coverage up to 1 kilometer, lower energy consumption, and up to 8191 stations. There are two types STAs in ah: sensor type to support sensor service and non-sensor type for fload service. In th research, it only focuses on non-sensor STA. For non-sensor STA, maximizing throughput more important than power consumption. Th research aims to see performance IEEE ah Enhanced Dtributed Channel Access (EDCA). To achieve that purpose, a mechanm needed to provide guarantees various services required by STA. EDCA an access mechanm used to set Quality Service (QoS) for IEEE standard through modifications in MAC layer. In th research, it focuses on one EDCA parameters, Arbitration Inter-Frame Space (AIFS). In addition, th research also focuses on ah feature Restricted Access Window (RAW) by changing RAW groups. From results research, it found that improvement scheme Arbitration Inter-Frame Space Number (AIFSN) AC BK = 2, AC BE = 1, AC VI = 1, AC VO = 1 has better performance compared to default scheme AIFSN AC BK = 7, AC BE = 3, AC VI = 2, AC VO = 2) an average throughput Mbps, average overall second and average PDR 62%. In addition, changes in RAW groups and RAW slots affect network performance. Th feature can improve throughput, average, and Packet Delivery Ratio. The goals th research to know effect AIFSN changes on AIFSN parameters, variation RAW group and RAW slot to throughput, average and packet delivery ratio. Index Terms AIFS, EDCA, IEEE ah, QoS, RAW I. INTRODUCTION INSTITUTE Electrical and Electronics Engineers (IEEE) ah new task group in IEEE Received: Aug. 16, 2017; received in reved form: Jan. 22, 2018; accepted: Jan. 23, 2018; available online: Apr. 25, It operates in sub-gigahertz band ( MHz in Europe; MHz in Japan; MHz, MHz, MHz in China; MHz in Korea and MHz in North America). It also can provide additional Wi- Fi network coverage compared to anor conventional Wi-Fi network that operates on 2.4 and 5 GHz band frequency [1]. In IEEE ah, energy consumption used lower than or WLAN standards. Th because each STA connected to AP will be in awake condition if STA receives or sends packets. Orwe, it will be in doze state, so it causes energy consumption in IEEE ah to be lower [2]. Although energy consumption on IEEE ah lower than exting WLAN standard, coverage area wider. It can reach up to 1 kilometer. Moreover, IEEE ah introduces an efficient paging and scheduling method to provide scalability for thousand stations. IEEE ah can serve up to 8191 stations [3]. IEEE ah also introduces a new mechanm, Restricted Access Window (RAW), which groups station (STA) into groups. Each group has limited access channel during a certain period, and th mechanm used to minimize collion [1]. There are two types STAs in IEEE ah, sensor type to support sensor service and non-sensor type for fload service. The sensor STA expected to have limited available power, low traffic volume and data frame small payload size. On or hand, for non-sensor STA, power consumption not critical and maximizing throughput more important [4]. Th research only focuses on non-sensor STA. A mechanm needed to provide guarantees services required by non-sensor STA [4]. IEEE e-2005 or e a standard for IEEE that sets service quality (QoS) for WLAN through modifications in Media Access Control (MAC) layer. IEEE e has two MAC protocols, D-

2 Cite th article as: A. Oktaviana, D. Perdana, and R. M. Negara Performance Analys On IEEE ah Standard Enhanced Dtributed Channel Access Mechanm, CommIT (Communication & Information Technology) Journal 12(1), 35 42, tributed Coordination Function (DCF) and Point Coordination Function (PCF). DCF based on Carrier Sense Multiple Access Collion Avoidance (CSMA / CA), whereas PCF a polling mechanm controlled by Access Point (AP). In its development, re a new protocol which called as Hybrid Coordination Function (HCF). In HCF, re are two access mechanms. HCF Controlled Channel Access (HCCA) centralized, and Enhanced Dtributed Channel Access (EDCA) dtributed [5]. However, se two MAC protocols are basically designed for data transmsion and do not guarantee QoS [6]. The research focuses on IEEE ah EDCA mechanm. Th mechanm much easier regarding implementation and analys, and most previously researchers focus on th mechanm [5]. In EDCA mechanm, re are three parameters to improve QoS, windows contents, Arbitration Inter-Frame Space (AIFS), and Transmsion Opportunity (TXOP). Th research focuses on AIFS parameters, where previous research [7] only focused on effect EDCA parameters on QoS WLAN IEEE e EDCF. It changes AIFSN in AIFS parameters so that it can improve QoS. The researchers take AIFSN in Ref. [7] that can improve QoS. In addition, th research focuses on how effect changes in RAW group to throughput,, and Packet Delivery Ratio (PDR). In previous research [8] for parameters tested only throughput and, but in th research for parameters tested are throughput,, and PDR. A. IEEE II. RESEARCH METHOD IEEE a standard implementation wireless that works on 2.4; 3.6; 5 and 60 GHz frequencies. All devices and provions that are wireless today follow IEEE standard. With th standard, it intended that every different wireless device can stay in touch different vendors. In IEEE , re are several task groups that have specifications frequency band, bandwidth, modulation scheme, channel architecture, maximum data rate, coverage, and maximum different transmit power [9]. Table I shows group on IEEE There are two major WLAN topologies. There ad-hoc and infrastructure. Ad-hoc WLAN peer to peer network that set to serve temporary need. There no central coordination ext in th topology. Meanwhile, infrastructure WLAN topology whose station will access wireless channel under coordination Base Station (BS) or AP [10]. Fig. 1. The use IEEE ah for IoT. Fig. 2. The RAW structure. B. IEEE AH IEEE ah a development IEEE It can be used to meet need Wireless Sensor Network (WSN) and Machine to Machine (M2M). IEEE ah also capable delivering smart solutions for smart metering, plan automation, e-health, and intelligent transport systems. In addition, IEEE ah can manage station (STA) in large quantities because it has a signaling hierarchy and extence power saving management. The main cases use IEEE ah are in industry, home automation, smart metering, plantation and agriculture, and health as illustrated in Fig. 1. In th application, mostly use wireless sensors by monitoring and working toger to pass collected data to network. IEEE ah introduces efficient paging and scheduling methods to provide scalability for thousands stations. Th only supports infrastructure mode and sends data at high speed on different conditions. Meanwhile, a better powersaving mechanm proposed to address high energy consumption on conventional Wi-Fi technology [1, 11]. C. Restricted Access Window (RAW) To support a devices associated AP, TGah has developed a new mechanm to reduce congestion in channel access. In th mechanm during a certain time, window called as Restricted Access Window (RAW). A group stations allocated a certain time slot used to access channel in RAW during a beacon interval as in Figure 2. The 36

3 Cite th article as: A. Oktaviana, D. Perdana, and R. M. Negara Performance Analys On IEEE ah Standard Enhanced Dtributed Channel Access Mechanm, CommIT (Communication & Information Technology) Journal 12(1), 35 42, TABLE I THE IEEE STANDARD Standard Freq Band BW Mod Scheme Channel Arc GHz 20 MHz BPSK to 256-QAM DSSS, FHSS b 2.4 GHz 21 MHz BPSK to 256-QAM CCK, DSSS a 5.0 GHz 22 MHz BPSK to 256-QAM OFDM g 2.4 GHz 23 MHz BPSK to 256-QAM DSSS, OFDM n 2.4, 5.0 GHz 24, 40 MHz BPSK to 256-QAM OFDM ac 5.0 GHz 20, 40, 80, 160 MHz BPSK to 256-QAM OFDM ad 60 GHz 2.16 GHz BPSK to 256-QAM SC, OFDM af MHz 6, 7, 8 MHz BPSK to 256-QAM SC, OFDM ah 900 MHz 1, 2, 4, 8, 16 MHz BPSK to 256-QAM SC, OFDM station not allowed to access channel when it outside RAW station slot. In addition, in beacon interval, re are several RAWs that are allowed to access channel [12]. In IEEE ah, each station uses two back-f states for EDCA to manage transmsions in and outside respective defined RAW. The first back-f state used outside RAW and second used in RAW slot. For first back-f state, station s back-f at beginning each RAW and returns it to continue back-f at end RAW. For second back-f situation, station initiates a back-f beginning a back-f state in RAW slot and removes back-f state at end RAW slot [8]. D. IEEE e IEEE e-2005 or e a standard for IEEE that governs QoS for WLAN through modifications to MAC layer. QoS designed to help end users become more productive by ensuring that users get reliable performance from networkbased applications. QoS refers to ability network to provide better service on certain network traffic through different technologies. IEEE e defines a new coordination function called as Hybrid Coordination Function (HCF). E. Enhanced Dtributed Channel Access (EDCA) EDCA designed to maximize QoS by adding functionality to DCF. Then on MAC layer, it defines four First in First out (FIFO) queues called as Access Category (AC). The access mechanm similar to DCF. However, it different for duration DIFS as it replaced by AIFS. Before entering MAC layer, each packet data received from above layer assigned to a specific user priority between 0 and 7. Th priority maps four AC such as background task traffic (AC BK), best effort traffic Fig. 3. The EDCA Parameters. (AC BE), video traffic (AC VI), and voice traffic (AC VO) [5]. There are three main parameters for QoS adjustment to EDCA describes in Fig. 3: (1) Arbitration Inter-Frame Space (AIFS): space-time after station sends packets, (2) Transmsion Opportunity (TXOP): period required by station to deliver packet, and (3) Contention Window (CW): back-f timer after AIFS and before station sends packets or frames. If medium idle, CW can be set to zero to be sent immediately. The CWs driven between CWmin and CWmax. EDCA provides a mechanm for dtinguhing data flow types so that th technique assumes that each type data stream has different priorities and each data stream will be classified under User Priorities (UP) and Access Category (AC). In total re are eight types User Priorities (UP) which are mapped into four types Access Category (AC) as in Table II. In ah system, an AP can indicate type STAs supports. It can select to support sensor type STA, non-sensor type STA, or both. In IEEE ah, each STA sensors and non-sensors access channel channels using different EDCA parameters as illustrated in Tables III and IV. For STA sensor access category best effort selected as default setting, while in STA nonsensor category access category voice access category as default. F. Simulation Model The simulation using Network Simulator 3.23 IEEE ah module. It modified by 37

4 Cite th article as: A. Oktaviana, D. Perdana, and R. M. Negara Performance Analys On IEEE ah Standard Enhanced Dtributed Channel Access Mechanm, CommIT (Communication & Information Technology) Journal 12(1), 35 42, TABLE II USER PRIORITY IN EDCA Priority User Priority (UP) Access Category (AC) AC Designation D Designation Lowest 1 AC BK Background BK 2 AC BK Background - 0 AC BE Best Effort BE 3 AC BE Best Effort EE 4 AC VI Video CL 5 AC VI Video VI 6 AC VO Voice VO Highest 7 AC VO Voice NC TABLE III EDCA PARAMETERS FOR SENSOR STA Access Category AIFSN CWmin CWmax Parameter TABLE V SIMULATION PARAMETERS. Information Background Physical Layer WLAN / IEEE Best Effort Transport Layer UDP Video Datarate 7.8 Mbps Voice Bandwidth 2 MHz Mobility Model Random Direction 2D Mobility Model STA Speed m/s TABLE IV Max range 100 m EDCA PARAMETERS FOR NON-SENSOR STA TX Current A > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION RX NUMBER Current (DOUBLE-CLICK HERE TO EDIT) A< Access Category AIFSN CWmin CWmax Idle Current A 4 Background Sleep Current A Best Effort consumption, Videoit uses TX, RX, 5Idle, and 7Sleep modes 15 in Space Number (AIFSN) in improvement scheme accordance Voice Ref. [7] found For in slot Ref. format [6] that 0, QoS maximum improves RAW compared slot count to default that scheme. usedthe to result calculate in Table duration 6. each slot 256. For slot format 1, maximum RAW slot count limit TheTable parameters 6. Parameters set out Variation in general AIFSN for Value Parameter Default Scheme Improvement Scheme ah network scenario can be seen in Table V. The simulation done through two scenarios. Each scenario presents a certain change. The first scenario presents a change in AIFSN to Number 30, 60, 90, 120 and 150 stations. Station The second scenario presents RAW group changes RAW Group to stations. 1 RAW Slot 1 1) Variation AIFSN Value Scenario Th scenario aims to simulate IEEE standard Fig. 4. Topology simulation. 2. EDCA Variation mechanm RAW Number and AIFS Group parameter Scenario Figure 4 Topology simulation Th scenario that AIFSN conducted to to change determine station. effect RAW addition RAW feature. The model a network group Thechanges on stations starts throughput, from 30average, and simulation an AP a stations PDR. addition The change 30 to in 150 stations and RAW maximum groups 1 and In simulation ah in NS3, re are two types dtributed using Uniform Dc Position Allocator as Number dtance Station between (N) /2. AP The and station stations 100 started slot format, 0 and 1. For slot format 0, maximum describe in Figure 4. The mobility user uses RandomDirection2dMobilityModel a range speed Coding Scheme (MCS) used in th scenario from 30 addition stations as many as 30 to 120 meters as describe in Table 7. The Modulation and RAW slots 64. For slot format 1, maximum limit for maximum dtance between AP and station 100 RAW slots 8. For slot format 0, maximum meters as describe in Table 7. In th scenario, it also uses about m/s. It simulates users. For energy RAW slot count that used to calculate duration each EDCA MCS parameters Mbps from improvement data rate and 2 scheme MHz bandwidth. The Arbitration Inter-Frame Space Number in first consumption, it uses TX, RX, Idle, and Sleep modes slot 256. For slot format 1, maximum RAW slot count scenario. in accordance Ref. [8]. (AIFSN) in improvement scheme limit The parameters set out in general for In simulation ah in NS3, re are foundtable in Ref. 7. Parameters [7] that Variation QoS improves RAW Group compared Number ah network scenario can be seen in Table 5. two types slot format, 0 and 1. For slot format 0, to default Parameters scheme. The result in Information Table VI. maximum RAW slots 64. For slot 2) Variation Number RAWGroup Number Group Scenario 1 and N/2 Th Table 5 Simulation parameters format 1, maximum limit for RAW slots scenario Number conducted Stations (N) to determine 30, 60, 90 effect and 120 Parameter Information Dtance 100 Meter Physical Layer WLAN / IEEE Number RAW Slot 1 Transport Layer UDP Datarate 7.8 Mbps III. RESULT AND ANALYSIS 38 Bandwidth 2 Mhz A. Variation AIFSN Value Result Mobility Model Random Direction 2D Mobility Model STA Speed 1.2 m/s 1.8 m/s AC BK BE VI VO BK BE VI VO CWmin CWmax AIFSN

5 es mber it for mum each ount r Each sents. The at tions d 100 ding bps rame RAW Group 1 when station 90 to 150. Th throughput reduction RAW Slot 1 most likely due to occurrence collions in channel. Th collion occurs because, in th scenario, 2. Variation RAW Number Group Scenario RAW groups and RAW slots used Th scenario Cite th article conducted as: A. Oktaviana, to determine D. Perdana, effect and1. R. Th M. RAW Negara means Performance re no channel Analysaccess On IEEE restriction ahfor each Standard Enhanced Dtributed Channel Access station, Mechanm, so all stations CommIT will (Communication try to compete & Information each or to group changes on throughput, average, and Technology) Journal 12(1), 35 42, access channel. Th causes throughput to PDR. The change in RAW groups decrease. 1 and Number Station (N) /2. The stations started from 30 addition TABLE stations VI as many as 30 to 120 maximum PARAMETERS dtance OF VARIATION between OF AIFSN AP and VALUE. station 100 meters as Parameter describe in Table Default7. Scheme In th Improvement scenario, Scheme it also uses EDCA parameters AC from BK improvement BE VI VO BKscheme BE VI in VO first scenario. CWmin CWmax AIFSN Number Station 30, 60, 90, 120 and 150 RAW Group 1 Table 7. Parameters Variation RAW Group Number RAW Slot 1 Parameters Information Number RAW Group 1 and N/2 TABLE VII Number Stations (N) PARAMETERS OF VARIATION RAW 30, 60, GROUP NUMBER. 90 and 120 Dtance 100 Meter Number Parameters RAW Slot Information 1 Fig. 6. Effect changes AIFSN to. Figure 6 Effect changes AIFSN to Number RAW Group 1 and N/2 Number III. RESULT Stations AND (N) 30, ANALYSIS 60, 90 and 120 In Figure 5, it can also be seen that in improved scheme, 6, it can throughput be seen that improved default com- scheme Dtance 100 m In Figure Number RAW Slot 1 A. Variation AIFSN Value Result performance pared to under default scheme, performance which improvement improvement scheme %. It based Inon improved average scheme, result highest addition throughput when stations. In default stations scheme, highest equal average to 60 or Mbps. when Then, lowest stations equal to when 150 like seconds. station The 150 or lowest Mbps when average station throughput 30, that overall about seconds average Mbps. total about second. In Figure The throughput 6, it can also improvement be seen that scheme has average better performance results compared improvement to scheme default schema. better Th than be-defaulcause, The inimprovement improvement obtained scheme, %. AIFSN In scheme. improvement smaller scheme, than highest default average scheme. It causes at AIFS to become smaller and medium access stations about 150 or seconds. priority to be higher. Higher priority can lead to Fig. Figure 5. Effect 5. Effect changes changes AIFSN AIFSN to throughput. Meanwhile, lowest when station to throughput higher chances accessing channel, which 30 or second average total reason why throughput can increase. On about In Figure RAW 5, it group can be changes seen onthat performance throughput, or hand, seconds. if AIFS greater, it can Figure default scheme average, under and PDR. performance The change in improvement reduce 6 also shows overallthat system throughput. increasing It because users will reviewed RAWbased groupson 1 andthroughput Number Station result (N) increase station s chance toaverage access. channel Th becomes because small. more scheme. It In addition, large AIFS s can also cause addition /2. The stations. stations In started default from scheme, 30 network to be under a large load. highest throughput addition when stations as many as 30stations to From Figure 5, it can be seen that throughput 120 maximum dtance between AP and equal to 60 or Mbps. Meanwhile, lowest decreases when station 90 to 150. Th station 100 meters as describe in Table 7. In when th scenario, stations it also uses 150 EDCA such parameters as from Mbps throughput reduction most likely due to occurrence collions in channel. Th collion average improvement scheme in overall first scenario. throughput around Mbps. occurs because, in th scenario, RAW groups and RAW slots used 1. Th In Figure 5, it can III. also RESULTS be seen ANDthat DISCUSSION in improved scheme, means re no channel access restriction for each throughput A. Variation AIFSN improved Value Result compared to default station, so all stations will try to compete each scheme, which In Figure 5, improvement it can be seen that %. performancein or to access channel. Th causes throughput default scheme under performance to decrease. improvement scheme. It reviewed based on In Figure 6, it can be seen that default scheme throughput result addition performance under performance improvement stations. In default scheme, highest throughput scheme. It based on average result when stations equal to 60 or addition stations. In default Mbps. Meanwhile, lowest when scheme, highest average when stations 150 such as Mbps stations equal to 150 like average overall throughput around seconds. The lowest when Mbps. station 30, that seconds average wit F sch obt imp val ove I sta bec cau cau cau B. F per sta thr , ove 39

6 ues can also cause network to be under a large load. rom Figure 5, it can be seen that throughput decreases en station 90 to 150. Th throughput reduction st likely due to occurrence collions in nnel. Th collion occurs because, in th scenario, scheme better than default scheme. The improvements ber RAW Cite groups th and article as: A. Oktaviana, RAW D. slots Perdana, used and R. obtained M. Negara by Performance improvement Analys scheme On are IEEE 8%. In ah scheme Th means re Standard no channel Enhanced access Dtributed restriction Channel for each Access improvement, Mechanm, CommIT highest (Communication PDR occurs & Information when tion, so all stations Technology) will try Journal to compete 12(1), 35 42, each or to stations equal to 30 i.e., 100%. Meanwhile, lowest ess channel. Th causes throughput to at station about 150 i.e., 12% rease. overall 62% PDR average. total about second. In Figure 6, it can also be seen that average performance improvement scheme better than default scheme. The improvement obtained %. In improvement scheme, highest average at stations about 150 or seconds. Meanwhile, lowest when station 30 or second average total about seconds. Figure 6 also shows that increasing > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER users will increase average. Th Figure because 6 Effect changes more stations AIFSN are trying to Fig. 7. Effect changes AIFSN to PDR. to access channel Figure 7. Effect changes AIFSN to PDR to transmit data packets. Therefore, each station has to wait for or stations to access channel. The length n Figure 6, it can be seen that default scheme In figure 7, it can be implied that more each waiting station causes to increase as formance under performance improvement stations on a network results in PDR to decline. Th stations increases. The average eme. It based on average result because more stations are trying to access channel. It in repair scheme has a better result compared ition stations. In default scheme, causes possibility collions to be larger. Thus, th can to scheme 1 (default scheme). Th as previously hest average described in when throughput results stations that cause packet to be a lot loss. The packet loss higher al to 150 like priority can seconds. cause The station s lowest chance to when causes PDR to decline. access channel station to be 30, higher. that The station does seconds not have to wait B. Variation RAW Group Number average total too long to access about medium. second. Therefore, repair n Figure 6, scheme it can also has better be seen average that average results compared Figure 8 shows that RAW Group = 1 under to formance improvement default scheme. scheme better than default performance RAW Group scheme = N/2. It based on eme. The improvement In Figure 7, obtained it can be seen %. that performance In throughput result addition rovement scheme, default highest scheme average under performance at stations. Fig. Figure 8. In Effect 8 Effect changes RAW changes RAW Group RAW Group Group scheme to= to throughput. 1, highest ber stations improvement about scheme. 150 or It based on seconds. throughput when stations equals 30 i.e., PDR results anwhile, lowest addition when stations. station In Figure 8, Mbps it can and also be lowest seen that RAW station Group = N / In 30 or default scheme, second highest average PDR total 90, 2, which throughput stations Mbps improves In RAW compared Group average to scheme throughput RAW = Group 1, = 1 highest scheme. throughput The increase %. when In RAW ue about stations seconds. overall about 1.04 Mbps. about to 30 i.e., 100%. Then, lowest Group = stations N / 2, equals highest 30 i.e., throughput Mbps and when igure 6 also shows that when increasing station users 150 will i.e., 24% lowest stations station equal to 12090, or which Mbp. rease average overall average. Th PDR because 58%. more FigureMeanwhile, 7 Mbps lowest average throughput at station overall about that shows that performance improvement scheme Mbps. Mbps average throughput overall about better than default scheme. The improvements obtained In Mbps. Fig. 8, it can also be seen that RAW Group by improvement scheme are 8%. In scheme The = N/2, RAW Group throughput = N / 2 scheme improves better compared than to RAW improvement, highest PDR occurs when Group RAW = 1 Group scheme. = 1 It scheme. indicates Thethat increase grouping %. stations can stations equal to 30 i.e., 100%. improve In RAW throughput Group s. = N/2, The highest throughput can be good because each station has limits to access channel, Meanwhile, lowest at station when stations equal to 120 or which limits purpose minimizing occurrence about 150 i.e., 12% overall 62% PDR Mbp. Meanwhile, lowest at contention and probability occurrence collions. average. station 60 that Mbps Then, contention and collion can cause network In Figure 7, it can be implied that more average throughput overall about Mbps. performance to drop. stations on a network results in PDR to decline. Th because more stations are trying to access channel. It causes possibility collions to be larger. Thus, th can cause packet to be a lot loss. The packet loss causes PDR to decline. B. Variation RAW Group Number Figure 8 shows that RAW Group = 1 under performance RAW Group scheme = N/2. It based on throughput result addition stations about to 30 i.e., 100%. Then, lowest when station 150 i.e., 24% overall average PDR 58%. Figure 7 shows that performance improvement The RAW Group = N/2 scheme better than RAW Group = 1 scheme. It indicates that grouping stations can improve throughput s. The throughput can be good because each station has limits to access channel, which limits purpose minimizing occurrence contention and probability occurrence collions. Then, contention and collion can cause network performance to drop. Figure 9 shows that increasing users will increase average. Th because more stations are trying to access channel to transmit data packets. Therefore, each station must Figure 9 Effect changes RAW Group to Figure 9 shows that increasing users will increase 40 average. Th because more stations are trying to access channel to transmit data packets. Therefore, each station must wait for or stations to finh accessing channel. The length each waiting group, each s stations that a Figure From Figu stations on a because more causes po cause pac causes PDR AIFSN val results improvem 1, AC_VO = scheme (AC_ The changes i network. throughput, a However, thi sta th research, well despite t RAW group. [1] Zhao, Y ah Engineeri Networki [2] Daneshfa Mechan

7 Mbps. causes possibility collions to be larger, so th can The RAW Group = N / 2 scheme better than RAW cause packet to decline. The packet loss Group = 1 scheme. It indicates that grouping stations can causes PDR to decline. improve throughput s. The throughput can be good because each station has limits to access channel, IV. CONCLUSION which Cite limits th article purpose as: A. Oktaviana, minimizing D. Perdana, occurrence and R. M. Negara AIFSN Performance change Analysaffects On IEEE network ah performance. From contention Standardand Enhanced probability Dtributed occurrence Channel Access collions. Mechanm, results CommIT (Communication research, it found & Information that AIFSN in Then, Technology) contention Journal and 12(1), collion 35 42, can cause network improvement scheme (AC_BK = 2, AC_BE = 1, AC_VI = performance to drop. 1, AC_VO = 1) has more performance compared to defaul scheme (AC_BK = 7, AC_BE = 3, AC_VI = 2, AC_VO = 2) = 2, The AC changes BE = 1, in AC VI = 1, AC RAW VO affect = 1) hasperformance o more performance network. The compared RAW mechanm to default can scheme increase o (AC BK throughput, = 7, ACaverage BE = 3,, AC VI PDR, = 2, and AC energy VO = consumption 2). The However, changesth in depends on RAW used affect evaluation metrics, performance stations, network. and The RAW traffic mechanm load on cannetwork. From increase th research, energy throughput, efficiency average mechanm, PDR, in ah work and energy well despite consumption. change However, AIFSN th depends onand IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < used 6 evaluation RAW group. metrics, stations, and traffic load on network. From th research, energy efficiency mechanm in ah works well despite group, each station must wait longer due to change AIFSN and REFERENCES RAW Figure 9 Effect changes RAW Group to stations Fig. 9. that Effect access changes channel. RAW Group to. group. [1] Zhao, Y. (2015). Analys Energy Efficiency in IEEE Figure 9 shows that increasing users will ah. AALTO UNIVERSITY, School Electrica REFERENCES increase average. Th because more Engineering, Department Communications and [1] Y. Zhao, stations are trying to access channel to transmit data Networking. Analys energy efficiency in ieee ah, Master s s, Masters Programme packets. Therefore, each station must wait for or stations to [2] Daneshfar, N. (2015). Performance Enhancemen in Communications Engineering, Aalto University, finh accessing channel. The length each waiting Mechanm IEEE ah Machine Communication station causes to increase along increasing System. [2] N. Daneshfar, Performance enhancement mechanm stations. [3] Park, ieee M. (2014) ah IEEE machine ah communication : Energy Efficient MAC The RAW Group Scheme = N/2 better than RAW system, Protocols Master s for s, Long Range Department Wireless Electronics [4] and IEEE Communications Standard Association, Engineering, Wireless Faculty LAN Medium LAN. IEEE. Group 1 scheme. It implies that grouping stations can hroughput improve average. The average can Computing Access and Control Electrical (MAC) Engineering, and Physical Tampere Layer (PHY be good for RAW Group N/2 because re are 30 stationsthat Specifications: Amendment 6: Sub 1 GHz License Group = N / University Technology, are divided into 15 groups, so a group contains 2 stations. The to RAW [3] M. Park, Exempt Ieee Operation, P802.11ah/D0.1, ah: Energy efficient May stations Fig. 10. in Effect one changes group will RAWcompete Group to access to PDR. channel, 7%. In RAW mac where waiting time for channels in idle condition less [5] Prasetya, protocolss. for (2015). long range Quality wireless Service lan, Improvement in Figure 10 Effect changes RAW Group to PDR when compared to 30 stations in one group. For 30 stations in one IEEE802.11e International EDCA Conference Scheme onusing Communications (ICC). IEEE, 2014, pp Enhanced Adaptive Mbp. From wait for Figure or10, stations it can tobe finh concluded accessing that channel. ber 60 that stations The length on a network each waiting cause station PDR causes to decline. Th to [4] IEEE standard for information technology overall about because increase more along stations are increasing trying to access stations. channel. It Telecommunications and information exchange causes The RAW possibility Group Scheme collions = N/2 to be better larger, than so th can between systems Local and metropolitan area han RAW cause packet to decline. The packet loss RAW Group 1 scheme. It implies that grouping networks Specific requirements Part 11: Wireless LAN Medium Access Control (MAC) and g stations can causes PDR to decline. stations can improve average. The ughput can be average can be good for RAW Group N/2 Physical Layer (PHY) specifications Amendment s channel, IV. CONCLUSION because re are 30 stationsthat are divided into 15 6: Wireless Access in Vehicular Environments, occurrence AIFSN change affects network performance. From collions. groups, so a group contains 2 stations. The 2 stations IEEE Working Group Std., results research, it found that AIFSN in network in one group will compete to access channel, where [5] S. Prasetya, B. Rahmat, and E. Susanto, Quality service improvement e edca improvement scheme (AC_BK = 2, AC_BE = 1, AC_VI = 1, AC_VO waiting = time 1) has for more channels performance in idle compared condition to less default scheme compared (AC_BK to 30= stations 7, AC_BE in one = 3, group. AC_VI For = 2, 30AC_VO stations scheme using enhanced adaptive contention window algorithm, in IEEE International Confer- = 2). The inchanges one group, in each station must RAW wait affect longer performance due to network. The stations RAW that mechanm access can channel. ence on Communication, Networks and Satellite increase throughput, From Figure average 10,, it can be PDR, concluded and energy that consumption. (COMNESTAT). IEEE, 2015, pp However, stations th on depends a network on cause used valuation PDR to decline. [6] R. Achary, V. Vaithiyanathan, P. Raj, and S. Nagarajan, Performance enhancement ieee metrics, Th because stations, more and stations traffic are load trying on to network. access From th channel. research, Itenergy causesefficiency possibility mechanm collions in ah to be le wlan by dynamic adaptive contention window, works well larger, despite so thchange can cause AIFSN packet toand decline. The in 16th International Conference on Advanced RAW group. packet loss causes PDR to Communication Technology (ICACT). IEEE, decline. o users will because more transmit data r stations to each waiting increasing han RAW stations can REFERENCES IV. CONCLUSION [1] Zhao, Y. (2015). Analys Energy Efficiency in IEEE ah. AIFSN AALTO change UNIVERSITY, affects network School performance. Electrical From Engineering, results Department research, it Communications found that and AIFSN Networking. in improvement scheme (AC BK [2] Daneshfar, N. (2015). Performance Enhancement Mechanm IEEE ah Machine Communication System. [3] Park, M. (2014). IEEE ah : Energy Efficient MAC 41 Protocols for Long Range Wireless LAN. IEEE. [4] IEEE Standard Association, Wireless LAN Medium 2014, pp [7] R. Qashi, M. Bogdan, and K. Hänssgen, Case study: The effect variable priority parameters on qos wlans ieee e edcf, in IEEE 3rd International Conference on Communication Stware and Networks (ICCSN). IEEE, 2011, pp

8 Cite th article as: A. Oktaviana, D. Perdana, and R. M. Negara Performance Analys On IEEE ah Standard Enhanced Dtributed Channel Access Mechanm, CommIT (Communication & Information Technology) Journal 12(1), 35 42, [8] L. Tian, S. Deronne, S. Latré, and J. Famaey, Implementation and validation an ieee ah module for ns-3, in Proceedings Workshop on ns-3. ACM, 2016, pp [9] S. S. B, A quantitative analys ah wireless standard, International Journal Latest Research in Engineering and Technology (IJLRET), vol. 2, no. 2, pp , [10] T. Anwar and W. G. P. Mui, Design and implementation a wireless network system in a smart campus, CommIT (Communication and Information Technology) Journal, vol. 1, no. 2, pp , [11] M. I. Denatama, D. Perdana, and R. M. Negara, Anal perbandingan kinerja protokol routing dsdv dan olsr untuk perubahan kecepatan mobilitas pada standar ieee ah, Jurnal Infotel, vol. 8, no. 2, pp , [12] M. Qutab-ud din, Enhancements and challenges in ieee ah - a sub-gigahertz wi-fi for iot applications, Master s s, Department Electronics and Communications Engineering, Faculty Computing and Electrical Engineering, Tampere University Technology,