IEEE ah Based M2M Networks Employing Virtual Grouping and Power Saving Methods

Transcription

1 2976 IEICE TRANS. COMMUN., VOL.E96 B, NO.12 DECEMBER 2013 PAPER Special Section on Network and System Technologies for Sustainable Society IEEE ah Based M2M Networks Employing Virtual Grouping and Power Saving Methods Kohei OGAWA a), Student Member, Masahiro MORIKURA, Fellow, Koji YAMAMOTO, Senior Member, and Tomoyuki SUGIHARA, Nonmember SUMMARY As a promising wireless access standard for machine-tomachine (M2M) networks, the IEEE task group ah has been discussing a new standard which is based on the wireless local area network (WLAN) standard. This new standard will support an enormous number of stations (STAs) such as 6,000 STAs. To mitigate degradation of the throughput and delay performance in WLANs that employ a carrier sense multiple access with collision avoidance (CSMA/CA) protocol, this paper proposes a virtual grouping method which exploits the random arbitration interframe space number scheme. This method complies with the CSMA/CA protocol, which employs distributed medium access control. Moreover, power saving is another important issue for M2M networks, where most STAs are operated by primary or secondary batteries. This paper proposes a new power saving method for the IEEE ah based M2M network employing the proposed virtual grouping method. With the proposed virtual grouping and power saving methods, the STAs can save their power by as much as 90% and maintain good throughput and delay performance. key words: M2M networks, wireless LAN, IEEE ah, EDCA, power saving, CSMA/CA 1. Introduction Machine-to-machine (M2M) networks, which are composed of power meters and various sensors, are currently attracting considerable attention for the realization of smart grids and smart communities. M2M networks are characterized by a large number of stations (STAs). A wireless medium access standard for M2M networks has been discussed by the IEEE task group ah (hereafter we write this abbreviated IEEE ah) [1], which is based on the cutting-edge wireless local area network (WLAN) standard such as the IEEE ac standard [2]. In IEEE ah, thousands of STAs are assumed to belong to an access point (AP). As the number of STAs increases, the throughput and delay performances of the system are significantly degraded when the carrier sense multiple access with collision avoidance (CSMA/CA) protocol is exploited. In normal cases, this problem does not occur because traffic in M2M networks supposed to be usually light. However, we should consider that temporary traffic congestion occurs. For example, such instances occur when the AP transmits a certain trigger frame to many STAs to collect STA information. Manuscript received March 13, Manuscript revised June 27, The authors are with the Graduate School of Informatics, Kyoto University, Kyoto-shi, Japan. The author is with the Allied Telesis Holdings K.K, Tokyo, Japan. a) ogawa@imc.cce.i.kyoto-u.ac.jp DOI: /transcom.E96.B.2976 Many STAs respond to the trigger frame and transmit their reply data frames. These events cause traffic congestion in an M2M network. To solve this problem, several STA grouping methods have been proposed, which can be classified into real and virtual grouping. The first method allocates all STAs into some groups and assigns a communication time period to each group according to a centralized control scheme of the AP to realize the real grouping method [3]. It needs overhead control information to support this scheme. The second method realizes the grouping by a distributed control scheme based on the CSMA/CA protocol. As a virtual grouping method, the distributed coordination function with virtual group (DCF/VG) scheme is proposed [4], [5]. The DCF/VG scheme achieves good throughput performance even when traffic congestion occurs; however, it suffers from a power consumption problem of STAs. According to the DCF/VG scheme, each STA always searches for the allocated slot, which is composed of a channel busy period and an idle period. An STA can transmit its data frame only at the allocated slot. Therefore, each STA always needs to sense channel busy states in order to detect assigned slot boundaries. To realize an M2M network composed of powerefficient STAs, this paper proposes the IEEE ah based M2M networks employing virtual grouping and power saving method. Virtual grouping is realized by the random arbitration interframe space (IFS) number (AIFSN) scheme to improve the throughput performance under the condition of many active STAs. An active STA means that it stays at an awake state to transmit data frames. The random AIFSN scheme is proposed for the precise quality of service (QoS) control of WLANs to transmit several types of data frames [6]. We apply this method to the M2M network to realize virtual grouping and improve the throughput performance of a system with 6,000 STAs. Moreover, we propose a novel power saving method using the virtual grouping method. Most STAs in the M2M networks are supposed to be battery-operated sensors. Therefore, power saving is a key issue. The legacy IEEE standard supports the power save poll (PS-Poll) method as a power saving method. In addition to the PS-Poll method, the IEEE n standard supports the power save multi-poll (PSMP) method [12]. Various power saving methods for wireless sensor networks and WLANs have been proposed [7] [11]. However, there are few schemes for single hop networks with a large num- Copyright c 2013 The Institute of Electronics, Information and Communication Engineers

2 OGAWA et al.: IEEE AH BASED M2M NETWORKS EMPLOYING VIRTUAL GROUPING AND POWER SAVING METHODS 2977 ber of STAs. Schemes proposed in [7] [9] are intended for multi-hop and ad-hoc mode systems. Those proposed in [10], [11] are single hop WLANs with a small number of STAs. For single-hop networks with a small number of STAs, the PS-Poll or PSMP methods are usually used to realize power saving. For power saving of the STAs, the IEEE ah adopts the PS-Poll method, which enables non-active STAs to sleep except during the period when they are receiving selected beacon frames. The PS-Poll method is effective under the usual light traffic conditions; however, it is ineffective under heavy traffic. An active STA that has data frames in its transmission queue cannot enter into the sleep state unless the queue becomes empty. Under heavy traffic conditions, STAs are likely to wait for a long period before the data frames are transmitted. Thus, the power consumption of STAs increases. In this case, the PSMP method, can conventionally be used. It is suitable for a small number of STAs, but not for M2M networks that have a large number of STAs, because the PSMP frame length becomes very long owing to the large number of STAs. For example, the PSMP frame length of 6,000 STAs is estimated to be 48 kb, which takes approximately 1.3 s to send and receive at a transmission rate of 300 kbit/s. The AP has to transmit and the STAs usually receive these extraordinarily long PSMP frames in case temporal traffic congestion occurs, which causes an increase in STAs power consumption. We, therefore, propose a new power saving method employing virtual grouping in addition to the legacy PS-Poll method. The proposed method enables the STAs to save on their power consumption in a distributed control manner. This paper is organized as follows. Section 2 describes the virtual grouping method with the random AIFSN scheme. The analysis of the virtual grouping is given in Sect. 3. Section 4 provides the description of the proposed power saving method employing virtual grouping. Section 5 evaluates the performance of the M2M network and discusses the effectiveness of the proposed power saving method. Finally, Sect. 6 presents the conclusion of this paper. 2. Virtual Grouping with Random AIFSN Scheme This section describes the random AIFSN scheme and virtual grouping. 2.1 Random AIFSN Scheme According to the random AIFSN scheme, an STA randomly selects its AIFSN value from the interval [AIFSN min, AIFSN max ], prior to every data frame transmission. AIFSN is an integral number, which decides the length of the AIFS as AIFS time = SIFS time + AIFSN slot time, (1) where slot time = 52 µs [1]. AIFS is a type of IFS, which is used in enhanced distributed channel access (EDCA) of Fig. 1 Medium access control with the random AIFSN scheme. The AIFSN value is randomly determined after the transmission. the IEEE e standard [13] to realize QoS control. The AIFS determined by Eq. (1) is expressed as AIFS(AIFSN). For example, AIFS(2) means that the AIFSN value is two. It is equal to the period of the distributed IFS. To explain the random AIFSN scheme the transmission procedure is shown in Fig. 1. In this figure, a cycle is composed of an idle channel period and the following busy channel period. According to the scheme, each STA waits for the AIFS time and backoff time before its transmission as well as the IEEE e channel access procedure. The transmission period is composed of a data frame period, a short IFS (SIFS) period, and an ACK frame period. Two cases could occur during the transmission period; successful data frame transmission and data frame collision. For the installation of this method, the parameter of AIFSN max has to be shared among the AP and the STAs. It is managed by the AP. The AP distributes it to each STA by a data frame which includes the parameter of AIFSN max to prepare for traffic congestion, when the STAs associate with the AP to start the data and control frame transaction. 2.2 Virtual Grouping In [6], the random AIFSN scheme was applied to a system with a small number of STAs to realize the precise QoS control for several types of data transmissions. We applied this scheme to M2M networks with a large number of STAs. As a result, we found that virtual grouping is feasible. The throughput performance is considerably improved by the virtual grouping method, which limits the number of contending STAs in the case of simultaneous data frame transmissions. We define two types of STAs. One is a contending STA that decreases its backoff counter value, and the other is a non-contending STA that does not decrease its backoff counter value during a cycle. The flowchart of the EDCA scheme is shown in Fig. 2. As shown in the flowchart, the STA that does not decrease backoff counter does not transmit any frames in the cycle. These events are shown as bold allows in Fig. 2. We call such STAs non-contending STAs. By this definition, the

3 2978 IEICE TRANS. COMMUN., VOL.E96 B, NO.12 DECEMBER 2013 Fig. 3 Contending and non-contending STAs with virtual grouping when random AIFSN scheme is employed. Fig. 2 Flowchart of the EDCA scheme. STAs are divided into two groups every cycle as follows: IF (AIFS N i > min 1 j N [AIFS N j + backo f f j ]) STA i is a non-contending STA; ELSE STA i is a contending STA; where STA i denotes the i-th STA, AIFS N j is the AIFSN value of STA j, backoff j is the backoff counter value of STA j and N is the total number of STAs. In Fig. 1, STA 1 decreases its backoff counter and detects the channel busy state at cycle 1. STA 2 detects the channel busy state before decreasing its backoff counter. STA 3 decreases its backoff counter to zero and then transmits its data frame. In this case, STA 1 and STA 3 may have transmitted a data frame during the cycle, but STA 2 cannot because it have not decreased its backoff counter. Thus, during the cycle, STA 1 and STA3 are considered as contending STAs, and STA 2 is considered as a non-contending STA. In this case, STAs are divided into two groups, namely, contending and non-contending STAs by the random AIFSN scheme. With this scheme, virtual grouping of the STAs is realized. Figure 3 shows the virtual grouping mechanism. After the data frame transmission, an STA obtains a new AIFSN value. If obtained AIFSN value is small, it is likely to be a contending STA. By contrast if obtained AIFSN value is large, it is likely to be a non-contending STA. Figure 4 shows the state transition diagram of the AIFSN distribution in the case of the random AIFSN scheme. For the sake of simplicity, AIFSN min and AIFSN max are selected as 2 and 4, respectively, in this figure. At the first state, each STA selects its AIFSN value from the interval [AIFSN min, AIFSN max ]. The number of STAs for each AIFSN value has a uniform distribution. Hereafter, we denote the STA of which the AIFSN value equals n as STA(AIFSN = n). We also denote the STA of which the AIFSN value is greater than n as STA(AIFSN > n). Then, the state is moved to the second state where the number of STA(AIFSN = 2)s decreases and becomes small, whereas the number of STA(AIFSN > 2)s increases. This transition is done by the following procedure: 1. STA(AIFSN = 2)s are likely to be contending to transmit their frames. 2. After their frame transmissions, they reset their AIFSN values from the interval [2, 4] with the random AIFSN scheme, which causes a decrease of the number of STA(AIFSN = 2)s and an increase of the number STA(AIFSN > 2)s. The similar transition occurs from the second state to the third state. Finally, most of the STAs have the AIFSN max value at the steady state. The majority become STA(AIFSN = AIFSN max )s which correspond approximately to the non-contending STAs, the minority, but a certain number, become STA(AIFSN < AIFSN max )s which correspond approximately to the contending STAs. We call this phase reduction phase at the steady state, since the number of STA(AIFSN < AIFSN max )s decreases in this phase. When the number of STA(AIFSN < AIFSN max )s becomes very small, STA(AIFSN = AIFSN max )s also contends. Some of them with small backoff counter values transmit their frames and reset their AIFSN values. The number of STA(AIFSN < AIFSN max ) increases. We call this phase provision phase.

4 OGAWA et al.: IEEE AH BASED M2M NETWORKS EMPLOYING VIRTUAL GROUPING AND POWER SAVING METHODS 2979 Table 1 IEEE TGah parameters. Parameters Value Channel band width 1 MHz Data rate 300 kbit/s MAC payload 200 B OFDM symbol 40 µs Slot time 52 µs SIFS 160 µs CWmin 15 CWmax 1023 Fig. 5 Average number of contending STAs with the random AIFSN scheme (AIFSN max = 20). Fig. 4 State transition diagram of AIFSN distribution. For the sake of simplicity, AIFSN max = 4 is selected in this figure. 3. Performance Evaluation and Analysis of Virtual Grouping The computer simulation results of virtual grouping are presented in this section. To evaluate the throughput, delay, and power consumption performance, we developed a simulation platform on a Monte Carlo simulation using C++ language, which emulates the IEEE protocol as closely as possible for real-world operation of each transmitting STA. The simulation parameters are shown in Table Evaluation of Virtual Grouping The evaluation of virtual grouping with the random AIFSN scheme is presented. A single-hop network, composed of an AP and many STAs, is considered according to the IEEE ah system model. For the sake of simplicity, a saturated traffic condition is assumed in which the transmission queues of all the STAs are always occupied by data frames. After a data frame transmission of an STA, the next data frame is immediately supplied to the STA [14]. Figure 5 shows the average number of contending STAs with the random AIFSN scheme in the steady state. The average number of contending STAs is kept very small with many active STAs. In this case, most of the STAs are usually at a non-contending state. When the number of STAs is equal to 6,000, the average number of contending STAs is only 16. It enables the M2M networks to improve the throughput performance drastically, which is operated with the CSMA/CA protocolunderheavy traffic conditions. Figure 6 shows the throughput performance under a saturated traffic condition versus the number of STAs. The value of AIFSN max is set 20 in order to maximize the throughput performance with 6,000 STAs. The transmission data rate equals 300 kbit/s. The throughput performance of the conventional DCF system decreases with an increase in the number of STAs. On the other hand, the virtual grouping method with the random AIFSN scheme drastically improves the throughput performance. When the STAs of different transmission data rates are mixed, the number of contending STAs is similarly limited; therefore, virtual grouping works effectively. The throughput performance is affectedby transmissionrates. However,

5 2980 IEICE TRANS. COMMUN., VOL.E96 B, NO.12 DECEMBER 2013 Fig. 6 Throughput performance under saturated traffic condition when all the STAs transmit data frames at 300 kbit/s. Fig. 8 Throughput performance under saturated traffic condition when 10% of the STAs have the error-prone links whose average frame error rate is 0.1, and the data transmission rate is 300 kbit/s. Fig. 7 Throughput performance under saturated traffic condition when half of the STAs transmit data frames at 300 kbit/s and the others at 150 kbit/s. the virtual grouping scheme by the medium access control employing EDCA with the random AIFSN scheme is not affected. Figure 7 shows the throughput performance when half of the STAs transmit at 300 kbit/s and the others at 150 kbit/s. By including low transmission rate STAs, the throughput performance becomes less than that shown in Fig. 6; however, it is improved similarly when compared with that of the conventional DCF system. Some of the STAs may have an error-prone link. We assume that 10% of the links are error-prone, of which frame error rate equals 0.1. Figure 8 shows the evaluation results of the total throughput performance. The difference in the throughput performances shown in Figs. 6 and 8 is small. The virtual grouping method effectively works in the case where some error-prone links are included. 3.2 Analysis of Virtual Grouping We described the mechanism of the virtual grouping method in Sect In this section, we analyze the mechanism of how virtual grouping is achieved, using computer simulation Fig. 9 AIFSN distribution of each state using the random AIFSN scheme (AIFSN max = 11). results. In this analysis, the value of AIFSN max is 11, which means that each STA chooses its AIFSN value as an integer number in the range of 2-11.We assume that the number of STAs is 10,000. Figure 9 shows AIFSN distribution of each state using the random AIFSN scheme. At the first state, all the STAs randomly choose their AIFSN values at the start of the computer simulation. Therefore, approximately 1,000 STAs have the same AIFSN value because of the range of AIFSN and the total number of STAs. The STA(AIFSN = 2)s immediately decrease their backoff counters and transmit the data frames. They reset their AIFSN values; therefore, the number of STA(AIFSN = 2)s decreases, and the number of STA(AIFSN > 2)s increases. If the number of STA(AIFSN = 2)s becomes small enough, STAs(AIFSN = 3) acquire a transmission priority. This is the second

6 OGAWA et al.: IEEE AH BASED M2M NETWORKS EMPLOYING VIRTUAL GROUPING AND POWER SAVING METHODS 2981 Fig. 10 Transition of the AIFSN distribution using the random AIFSN scheme (AIFSN max = 11). Table 2 Example of the steady distribution of AIFS N. AIFSN Cycle , ,775 62, ,775 62, ,775 62, ,775 62, ,776 62, ,324 62, ,323 62, ,323 62, ,323 62, ,323 state. As in the case of STA(AIFSN = 2)s, the number of STA(AIFSN = 3)s decreases, and the number of STA(AIFSN > 3)s increases. Then, STA(AIFSN = 4)s follow them. Finally, most STAs become STA(AIFSN = AIFSN max )s at the steady state. This evolution process is shown in Fig. 10. The number of STA(AIFSN < AIFSN max )s decreases with a passage of cycles. The number of STAs with large AIFSN values decreases gradually because they may choose smaller AIFSN values after their transmissions, and these STAs with smaller AIFSN values acquire the transmission priority. Then, they will have larger AIFSN values due to the reset of their AIFSN values. Table 2 shows examples of AIFSN distribution in the steady state. Most STAs are STA(AIFSN = AIFSN max )s at every cycle. After the 62,149th cycle, STA(AIFSN = AIFSN max )s contend, and some of them transmit their data frames. In this case, a collision event occurs because more than one STAs transmit their data frames. Then, STAs(AIFSN < AIFSN max )are newly provided. There are two phases in the steady state as described in Sect The reduction phase distribution allows the minor STAs that have smaller AIFSN values to contend. The number of contending STAs at the steady state is shown in Fig. 11. The number of contending STAs stays small except at some discrete time instances. These Fig. 11 Transition of the number of contending STAs at the steady state. The number of STAs is 10,000. instances correspond to the provision phase in which STA(AIFSN < AIFSN max )s are provided, such as the 62,150th cycle in Table 2. After the provision phases, the number of contending STAs immediately gets small. The distribution of the reduction phase limits the number of contending STAs where most STAs are STA(AIFSN = AIFSN max )s, and there are some STAs with smaller AIFSN values. They prevent the majorities from contending. 4. Proposed Power Saving Method Employing Virtual Grouping We proposes the power saving method which makes noncontending STAs, a large majority, to sleep in addition to the PS-Poll method. The conventional power saving methods work well under light traffic conditions with a small number of STAs. However, under traffic congestion conditions with a large number of STAs, these methods degrade the throughput and power consumption performance. The objective of the proposed method is to reduce the power consumption of the system under traffic congestion conditions. The key idea of the proposed method is to make the non-contending STAs sleep. The flowchart of this scheme is shown in Fig. 12 for active STAs that have data frames in their transmission queues. First of all, they undergo a virtual grouping process with the random AIFSN scheme to transmit the data frame as explained in Sect. 2. During the virtual grouping process, each active STA experiences either of the states the contending or the non-contending state at every cycle. Each STA counts the number of successive non-contending states and enters into the sleep state if the number exceeds a threshold r th. After sleeping for a certain sleep period T sleep, the STA returns to the active state. When an active STA decreases its backoff counter value, it stays in the active state and tries to transmit its data frame. Each STA repeats this procedure until its queue becomes empty. The proposed method realizes the power saving method in a distributed control manner. With a small value of r th, it is easy for the STAs to sleep, although they are likely to lose opportunities to transmit their data frames. The

7 2982 IEICE TRANS. COMMUN., VOL.E96 B, NO.12 DECEMBER 2013 Fig. 13 Time chart of the downlink traffic channel access with the PS-Poll method. Table 3 Power consumption parameters [15]. State Transmitting Receiving Data processing Channel sensing Sleep Value 340 mw 260 mw 56 mw 204 mw 0.19 mw Fig. 12 Flowchart of the proposed power saving method for active STAs. 5. Evaluation of the Proposed Power Saving Method The computer simulation results of the proposed power saving method are presented in this section. The simulation parameters are shown in Tables 1 and 3. In the sleep state in Table 3, some parts of the circuit are awake and the association with the AP is maintained [15]. T sleep value is restricted by the IEEE standard and the characteristics of wireless sensor devices. With regard to the IEEE standard, the STAs have to receive beacon frames which include delivery traffic indication maps (DTIMs). Hence, T sleep must be smaller than the beacon frame interval. The beacon interval of IEEE based WLANs is usually 100 ms. In the IEEE ah standard, 1/10 down clocking is considered. Therefore we assume 1 s as the beacon interval. In order to receive the beacon frames from the AP, the STAs must wake up within 1 s. On the other hand, T sleep must be greater than one cycle period, the average of which is approximately 3 ms. As a time unit for the sleep mode, we select 100 ms as T sleep. In this proposed power saving method, data transmissions with the PS-Poll method are assumed. In this case, the STAs are usually in a sleep mode to save power consumption except for receiving the DTIMs that are included in beacon frames from the AP. After receiving the DTIMs, the STAs know existence of their downlink data frames at the AP. When an STA wakes up to request the downlink data frame, it transmits a PS-Poll frame to the AP. Then, the AP transmits the data frame to the STA a SIFS period after successful reception of the PS-Poll frame. This time chart is shown in Fig. 13. Therefore, only the STAs perform the channel access procedure for the downlink data frames. The proposed method effectively works when the AP transmits data frames to the STAs. 5.1 System Description A single-hop network with an AP and 6,000 STAs is considered according to the IEEE ah system model. Each STA is assumed to have one data frame in its transmission queue at the start of the simulation to evaluate the temporary uplink trafficcongestionstate. Inthiscase, asthesimulation time progresses, the traffic congestion is mitigated owing to an increase of the number of non-active STAs that have successfully transmitted their data frames. For the sake of simplicity, the AP transmits no data frames or beacon frames during this period. The system performance is evaluated using the two criteria. The first is the maximum delay, which is the required time for all the STAs to transmit the data frames. The second one is the consumed energy, which is the amount of energy that all the STAs consumes within this period. 5.2 Simulation Results We shows the improved performance of the system with the proposed method and derive an optimum value for r th. In this subsection, T sleep is fixed at 100 ms. Figure 14 shows the consumed energy of the 6,000 STAs versus r th. The conventional method indicates the random AIFSN scheme with the PS-Poll method for power saving. This figure shows that the proposed method reduces the consumed energy by 96% as compared with the conventional method at r th = 1. The consumed energy uniformly

8 OGAWA et al.: IEEE AH BASED M2M NETWORKS EMPLOYING VIRTUAL GROUPING AND POWER SAVING METHODS 2983 Fig. 14 Total consumed energy of the STAs versus r th. r th is the sleep threshold. Fig. 16 Total consumed energy of the STAs versus r th. r th is the sleep threshold. The number of active STAs is 100. Fig. 15 Maximum delay time for 6,000 STAs to transmit all data frames versus r th. r th is the sleep threshold. Fig. 17 Maximum delay time for 100 STAs to transmit all data frames versus r th. r th is the sleep threshold. increases with value of r th, as the STAs are unlikely to sleep with a large r th value. Considering only the consumed energy of the STAs, r th = 1 is the best value. Moreover, we should consider the delay performance when choosing the r th value. Figure 15 shows the maximum delay time for the 6,000 STAs to transmit versus r th. The maximum delay decreases with an increase in the r th value and converges to the conventional value. Assuming that the permissible increase in the maximum delay is less than 5%, r th should be greater than 1. Considering the foregoing argument, r th = 2 is the optimal value. Using this value in Fig. 14 the consumed energy is reduced to 4.6 kj, which is 10% of the conventional method. We evaluate the total consumed energy of the STAs versus r th where the number of active STAs equals 100, and the maximum delay time for 100 active STAs to transmit all the data frames versus r th as shown in Figs. 16 and 17, respectively. These simulation results show that r th = 2isthe optimal value not only in the heavy traffic condition but also in the light traffic condition. Finally, we evaluate the consumed energy of the pro- Fig. 18 Total consumed energy of the STAs versus number of active STAs. posed method at different conditions where some STAs are active and the others are not. Figure 18 shows the consumed energy versus the number of active STAs. In this

9 2984 IEICE TRANS. COMMUN., VOL.E96 B, NO.12 DECEMBER 2013 References Fig. 19 Maximum delay time for all the STAs to transmit the data frames versus number of active STAs. case, r th = 2 is selected according to the discussion previously mentioned. The proposed method reduces the consumed energy of the STAs in all cases. On the other hand, the maximum delay time for all the active STAs to finish their transmissions is increased by less than 5%, as shown in Fig. 19. These results show that the proposed power saving method can be applied to various conditions without parameter adaptation. For the conventional PS-Poll method, an STA requests a downlink data frames by using a PS-Poll frame when the STA wakes up. The PS-Poll method itself is common to both the proposed method and the conventional method. If we assume the downlink data frames, there will be the similar increase in maximum delay for the proposed method and the conventional method. 6. Conclusion To realize a good throughput and power-efficient M2M network employing the CSMA/CAprotocol, this paper has proposed the IEEE ah based M2M networks employing virtual grouping and power saving methods. To solve the throughput performance degradation caused by contention of many STAs, this paper has shown that virtual grouping with the random AIFSN scheme achieved a good throughput performance. To reduce the power consumption of the STAs when the CSMA/CA protocol is used, this paper proposed the power saving method combined with virtual grouping. It enabled power saving in a distributed control manner. The simulation results show that the proposed method reduced the energy consumption of the STAs by 90% within a 5% increase in the delay, compared with the system using the virtual grouping method and the conventional PS-Poll method. [1] IEEE P task group ah Meeting update, ieee802.org/11/reports/tgah update.htm [2] IEEE P802.11ac/D2.2 Draft Std. Amendment 4: Enhancements for Very High Throughput for Operation in Bands below 6 GHz, [3] M. Park, et al., RAW slot assignment, IEEE /1321r0, Nov [4] S.M. Kim and Y.J. Cho, Performance evaluation of IEEE distributed coordination function with virtual group, IEICE Trans. Commun., vol.e88-b, no.12, pp , Dec [5] S.M. Kim and Y.J. Cho, A virtual grouping scheme for improving the performance of IEEE distributed coordination function, Proc. Wireless Networks and Emerging Technologies, July [6] S. Gaur and T. Cooklev, Introducing finer prioritization in EDCA using random AIFSN, Proc. TridentCom 2007, Orlando, Florida, USA, May [7] C. fan Hsin and M. Liu, Network coverage using low duty-cycled sensors: Random & coordinated sleep algorithms, Proc. 3rd International Symposium on Information Processing in Sensor Networks, pp , [8] J. Liu, X. Jiang, S. Horiguchi, and T.T. Lee, Analysis of random sleep scheme for wireless sensor networks, Int. J. Sensor Networks, vol.7, no.1, pp.71 84, [9] Y.S. Chen, M.K. Tsai, L.S. Chiang, and D.J. Deng, Adaptive traffic-aware power-saving protocol for IEEE ad hoc networks, Proc. IEEE 17th International Conference on Parallel and Distributed Systems, pp , Dec [10] D. Qiao and K.G. Shin, Smart power-saving mode for IEEE wireless LANs, Proc. IEEE INFOCOM, pp , March [11] X. Hu, Z. Chen, and Z. Yang, Energy-efficient scheduling strategies in IEEE wireless LANs, Proc. IEEE Computer Science and Automation Engineering, pp , May [12] IEEE Std n Amendment 5: Enhancements for Higher Throughput, [13] IEEE Std e Amendment 8: Medium Access Control (MAC) Quality of Service Enhancements, [14] G. Bianchi, Performance analysis of IEEE distributed coordination function, IEEE J. Sel. Areas Commun., vol.18, no.3, pp , March [15] D.M. Dobkin and B. Aboussouan, Low power Wi-Fi (IEEE ) for IP smart objects, GainSpan Corporation, Kohei Ogawa received his B.E. degree from Kyoto University, Japan, in He is currently an M.E. student at Kyoto University, Japan. He has been engaged in the research of the protocol for Wireless M2M networks. Acknowledgment This work was supported in part by Grant-in-Aid for Scientific Research (B) (no ).

10 OGAWA et al.: IEEE AH BASED M2M NETWORKS EMPLOYING VIRTUAL GROUPING AND POWER SAVING METHODS 2985 Masahiro Morikura received his B.E., M.E., and Ph.D. degrees in electronics engineering from Kyoto University, Kyoto, Japan in 1979, 1981 and 1991, respectively. He joined NTT in 1981, where he was engaged in the research and development of TDMA equipment for satellite communications. From 1988 to 1989, he was with the Communications Research Centre, Canada, as a guest scientist. From 1997 to 2002, he was active in the standardization of the IEEE a based wireless LAN. He received the Paper Award and the Achievement Award from IEICE in 2000 and 2006, respectively. He also received the Education, Culture, Sports, Science and Technology Minister Award in 2007 and Maejima Award in Dr. Morikura is now a professor in the Graduate School of Informatics, Kyoto University. He is a member of the IEEE. Koji Yamamoto received the B.E. degree in electrical and electronic engineering from Kyoto University in 2002, and the M.E. and Ph.D. degrees in informatics from Kyoto University in 2004 and 2005, respectively. From 2004 to 2005, he was a research fellow of the Japan Society for the Promotion of Science (JSPS). Since 2005, he has been with the Graduate School of Informatics, Kyoto University, where he is currently an associate professor. From 2008 to 2009, he was a visiting researcher at Wireless@KTH, Royal Institute of Technology (KTH) in Sweden. His research interests include game theory, spectrum sharing, and M2M networks. He received the PIMRC 2004 Best Student Paper Award in 2004, the Ericsson Young Scientist Award in 2006, and the Young Researcher s Award from the IEICE of Japan in He is a member of the IEEE. Tomoyuki Sugihara received his B.P. degree from Kyoto Sangyo University, Japan, in He is currently a director of Allied Telesis Holdings, K.K., Japan. He has been engaged in the research & development of Ethernet and IP network products and protocols.