RSP-WRAN: Resource sharing protocol for inter/intra WRAN communications

Abstract

IEEE 802.22 Wireless Regional Area Network (WRAN) standard employs cognitive radio techniques to utilize the white (i.e., unused) spectrum in the TV frequency band. IEEE 802.22 WRAN employs cellular topology (i.e., point to multi-point communications). IEEE 802.22b is a modified version of IEEE 802.22 that allows peer-to-peer communications among WRAN users of the same cell. One of the issues that significantly affects the performance of WRAN is the spectrum sharing between overlapped WRAN cells (i.e., inter spectrum sharing) as well as among the users of the same cell (i.e., intra spectrum sharing). In this paper, we propose Resource Sharing Protocol for Inter/Intra WRAN communications (RSP-WRAN), which is based on maximizing the Jain’s fairness index in the spectrum-sharing process. Simulation results show that RSP-WRAN, compared to other protocols, achieves high performance in terms of fairness and network throughput.

Keywords

Resource sharing protocol WRAN users Jain’s fairness index spectrum-sharing

1. Introduction

Due to the continuous development of multiple wireless applications [1,7,10,18,19,21,24,30,31,65] and services [3,4,6,11,17,22,25–27,33,50,66], spectrum scarcity is becoming a serious problem [5,8,16,20,23,28,29,32,42,43]. Moreover, a thorough study and analysis of spectrum usage that was conducted by the Federal Communications Commission (FCC) demonstrated that the spectrum is inefficiently utilized in several geographical areas [54]. For example, in New York City, according to the FCC study, the spectrum usage is only about 13.1% for the spectrum between 30 MHz and 3 GHz. The problem is resulting from the static channel assignment policy for the spectrum; many users are idle for long time in their specific licensed bands of spectrum and no other users can use such bands, resulting in inefficient usage of the spectrum.

To mitigate this problem and increase the spectrum utilization, the FCC in 2003, proposed cognitive radio (CR) technology [51,55]. This technology allows unlicensed users (called secondary users) to use the licensed bands of spectrum during the periods when the licensed users (called primary users) are not using these bands. These bands are called spectrum holes or white spaces. According to this technology, the secondary users should not cause harmful interference to the primary users. CR technology has emerged as a solution to spectrum scarcity; it enables efficient and flexible spectrum usage [61]. The first standard for cognitive radio networks was the IEEE 802.22 WRAN standard [14,15,37,47,52,69]. IEEE 802.22 working group includes several known corporations (e.g., Intel, ST Micro, Motorola, CRC, Samsung, Philips, and Nokia). WRANs employ cognitive radio techniques to opportunistically utilize the white spaces in the TV frequency spectrum (VHF/UHF: 54 to 862 MHz) without harmful interference between the WRAN users, which are called customer premise equipment (CPE), and licensed TV users. In particular, the CPEs need to sense TV spectrum to detect the spectrum holes. Once a TV primary user is detected in any TV band, the CPEs must immediately switch to another available band or stop their transmission to avoid any interference. The first applied topology for WRAN is cellular topology (i.e., point to multi-point communications) as shown in Fig. 1. The coverage radius of each WRAN cell is about 33 km and may reach up to 100 km. In each WRAN cell, the CPEs communicate with their base station (BS) over fallow TV broadcast bands (i.e., not used by TV users). The BSs and CPEs have the ability to switch between channels without transmission errors and adapt their sensing, bandwidth, coding, and modulation schemes. Orthogonal frequency division multiple access (OFDMA) is used in up/down links. The WRAN standard did not specify any spectrum-sensing technique [52]. Different sensing techniques proposed for cognitive radio technology can be employed in WRAN [52,62]. The main application of WRAN is supporting remote and rural areas with wireless broadband access [56]. In USA, Africa and Asia, a significant percentage of the population exist in such regions [56]. Other applications of WRAN include multi-dwelling units, single-family residential, public and private campuses, small businesses, multi-tenant buildings, etc. More information about WRAN can be found in [34,53,70].

1.1. Problem statement

One of the major issues in WRAN communications is finding a strategy that allows the WRAN users to efficiently and fairly share the available spectrum (i.e., unused by TV stations) without causing interference on each other. This problem is called coexistence problem [41] and it could be between the CPEs of one WRAN cell or between geographically overlapped WRAN cells (see Fig. 1) and it is called in this case self-coexistence problem.

Fig. 1.

Example of overlapped WRAN cells.

For the self-coexistence problem, if each cell of the overlapped WRAN cells, after the sensing process, randomly use (i.e., without any coordination with other cells) any one of the available channels, then it is likely that an interference between these cells will occur. This kind of coexistence is not easy to solve due to different reasons. Mainly, the coverage range of a WRAN cell can reach up to 100 km, which means a wide interference range; greater than that of other unlicensed technologies. Accordingly, extensive MAC control messaging will be required, which affects spectrum utilization. Moreover, the wide coverage range results in a long round-trip delay that makes the coordination more difficult. Furthermore, for WRAN, there is no licensed spectrum for control signaling which also increases the difficulty of the problem. Unlike other IEEE 802 standards, the IEEE 802.22 working group requires that the self-coexistence problem should be considered as part of the standard and not after issuing the standard [15]. Researchers in this field proposed several techniques and schemes to solve or mitigate self-coexistence problem. Some of them will be introduced later in Section 2.

For the coexistence problem between the CPEs of one cell, the original standard for WRAN (i.e., IEEE 802.22 [15]) solves this problem by adopting a centralized topology for WRAN. According to this topology, the BS in a cell works as a relay for all communications between the CPEs inside the cell. In other words, in this structure, there is no direct communication between any two CPEs in the same cell. Hence, the BS will control and eliminate any interference between communications inside the WRAN cell. However, this is inefficient solution. It results in a limited network capacity because each down-link channel is reserved for one time slot to be used by only one CPE. In other words, in this solution, there is no frequency reuse. Accordingly, a new WRAN working group called IEEE 802.22b was established to enhance the network capacity by allowing the direct CPE to CPE communication inside one cell [46]. Different techniques and algorithms have been proposed to coordinate and manage the interference between different peer-to-peer communications inside a WRAN cell. Some of them are discussed later in Section 2.

The two coexistence problems can be treated as a resource sharing problem. The method used for resource sharing in a network has a severe effect on the network performance in terms of different metrics such as network throughput, fairness, and interference. Due to the randomness in the availability of channels in cognitive radio networks, the spectrum sharing issue is becoming more difficult. Moreover, in WRANs, unlike other cognitive radio networks, due to the wide coverage region, each WRAN cell (or CPE) could have different number of available channels as well as different requests (i.e., the amount of data to be transmitted). Therefore, the spectrum sharing issue in WRAN is more difficult to solve than other networks.

Fig. 2.

Example of two beacon groups.

1.2. Contributions

In this paper, we study and consider the two coexistence problems. We propose a protocol that addresses the spectrum sharing among different overlapped WRAN cells and among the users of the same cell. The proposed protocol is called Resource Sharing Protocol for Inter/Intra WRAN communications (RSP-WRAN). The design of RSP-WRAN aims at improving the network performance in terms of network throughput, fairness, and interference. RSP-WRAN allows the WRAN users (cells and CPEs) to efficiently and fairly share the available spectrum without causing interference on each other. The operation of RSP-WRAN can be summarized as follows. Firstly, each group of overlapped cells called beacon group BG (see Fig. 2) selects one BS as a coordinator for that group. Then each CPE in each cell sends to its BS information about its location, the list of its available channels, and its request (i.e., the amount of data to be sent). These are forwarded by BSs to their coordinators. Then, each coordinator in a BG builds up two tables: overlay table that depicts the interference map of the overlapped CPEs in the BG and channel table that shows the available channels for each CPE in the BG. After that, each coordinator runs a spectrum sharing algorithm that allows CPEs in BG to use the available channels efficiently, fairly, and without interference among them. The proposed algorithm increases the spectrum capacity by allowing channel reusing among the non-overlapped CPEs. Moreover, the algorithm selects the most fairly channel assignment among all possible assignments. Since CPEs may have different demands (i.e., a number of channels that satisfy its transmission), the proposed algorithm defines a parameter for each CPE to measure the closeness of its demand and the channels assigned to it. Then, Jain’s fairness index [48,49] is calculated for these parameters. Finally, the assignment that results in maximum Jain’s fairness index is selected. RSP-WRAN will be thoroughly explained and discussed in Section 3. Our simulation results show that the proposed RSP-WRAN achieves higher network performance, in terms of fairness and throughput, compared to other proposed protocols.

1.3. Organization

The rest of the paper is organized as follows: Section 2 presents some related works. In Section 3, our proposed protocol RSP-WRAN is discussed and illustrated. In Section 4, we use simulations to evaluate the performance of RSP-WRAN. Finally, our concluding remarks are drawn in Section 5.

2. Related work

Several techniques have been proposed to deal with the coexistence problem in WRAN communications. Some of them were specified for inter cell communications (i.e., the coexistence of overlapped WRAN cells), others were proposed for intra cell communications (i.e., the coexistence of CPEs in one cell), and others considered both inter and intra cell communications.

Fig. 3.

IEEE 802.22 MAC self-coexistence schemes.

To mitigate the coexistence problem among overlapped WRAN cells, a certain coexistence approach was defined in IEEE 802.22 standard [13], see Fig. 3. This approach consists of four processes: spectrum etiquette, interference-free scheduling, dynamic resource renting and offering (DRRO), and adaptive on demand channel contention (AODCC). For spectrum etiquette, each WRAN base-station searches for available channels that not used by its neighbor cells. In the case of no available channels, the base station goes to the next process (i.e., interference-free scheduling) to share the same channels with its neighbors in a non-interfering approach. If the owner of the channel does not allow the channel sharing process, the base station must run the DRRO process. In this process, the cell that has free channels is called offeror and the one that requires free channels is called renter. The offeror broadcasts its available offered channels and a required minimum number of credit tokens (MNCT). Each cell has a certain credit token budget which is similar to money, more information can be found in [39,40]. If the renters hear the broadcasted information, they reply by sending renting requests which consists of the required channels and the credit tokens that can be paid. After that, the renters with the higher credit tokens is granted the required channels. Then, if the base station still needs more channels, the base station runs the final process (i.e., AODCC). In this process, the base station sends to the channel owner a contention request includes the credit tokens that can be paid. If the credit tokens is higher than the MNCT of the channel owner, the channel owner releases the requested channels; otherwise, it rejects the request.

In [38], the authors proposed a contention-based approach for inter cell communications. According to this approach certain messages (i.e., channel contention request, channel contention reply, and channel contention acknowledgement) must be exchanged at each channel reservation time. This leads to a high delay because the wide range of WRAN cell (i.e., 100 km) results in a round-trip propagation delay of about 300 μsec. In [9,67], several techniques were proposed to improve the dynamic-frequency-hopping (DFH) scheme [13] with the aim of mitigating the coexistence problem between overlapped WRAN cells. However, some of these techniques need a high signaling overhead and others require specific hardware such as multiple transceivers and directional antennas. The authors in [39] proposed a protocol for channel allocation in a multiple overlapped WRAN cells scenario. This protocol is called a credit token based rental protocol and it is based on the concept of ascending auctions among the overlapped cells, which can be applied before the DRRO stage in the coexistence mechanism shown in Fig. 3. In [36], the authors proposed a channel sharing technique for overlapped WRAN cells. This technique is based on a credit token using auctioning. In this technique, the authors modeled DRRO as a progressive second price auction to make sure that the overlapped cells will determine their channel requests in a truthful manner. In [12], the authors proposed a Vickrey-based auction mechanism for channel sharing between overlapped cells. This mechanism is a kind of sealed-bid auction that needs the renters to firstly bid and then they will receive the bidding results only one time. Accordingly, the auction is completed in a fast way, resulting in more utilization for the available TV channels and reduced the signaling overhead. The authors in [57] established a sealed-bid double auction between overlapped TV broadcasters and WRAN service providers. Also, in [57], the authors proposed a pricing approach for channel sharing among the WRAN base stations and their users. The authors in [52] proposed a resource transaction algorithm that is based on DRRO and AODCC processes.

It is worth mentioning that all presented solutions in [12,36,39,57], and [52], only consider resource transactions but not interference-avoidance channel selection and spectrum etiquette, which were considered in [63]. The authors in [63] used game theory to analyze the coexistence problem among overlapped WRAN cells. They employed a modified minority game to analyze and model the problem and they proposed a mixed strategy that should be followed by the overlapped cells to achieve the Nash equilibrium. For the cost function in [63], the authors used a function of the form $M N / (M - N)$ , where M is the number of available channels and N is the number of overlapped WRAN cells. This cost function is intuitively selected, which means it might lead to a non-optimal solution and may be unsuitable in some cases, for example when $M = N$ . In [58], the authors proposed an approach for resource sharing among overlapped WRAN cells. According to this approach, each group of overlapped cells elect a coordinator that collects information from the overlapped cells and then runs a certain algorithm that allows the overlapped cells to fairly share the available channels. Based on the results of the resource sharing algorithm, the coordinator broadcasts the transmission map to all overlapped cells. In [2], the authors considered the problem of coexistence between overlapped WRAN cells. The authors proposed a distributed approach based on a discrete-time Markov chain. They modeled the problem as an absorbing discrete-time Markov chain. In this model, they defined the absorbing state as the state where all the overlapped cells select different available bands. Finally, the authors solved for the optimal switching probability that results in having a minimum absorbing time. In [44], the authors proposed an On Demand Spectrum Contention (ODSC) protocol, where the overlapped WRAN cells run a contention mechanism in a distributed manner (i.e., without a central coordinator). Based on the results of the contention mechanism the cells access the available channels.

In [60], the authors proposed a resource sharing algorithm for intra-cell WRAN communications (i.e., among the CPEs of the same cell). This algorithm is called peer-to-peer algorithm (P2P) and it is based on channel reuse to achieve the fairness and network capacity maximization. In [59], the authors proposed an inter/intra resource sharing algorithm, called 2I-RSA. The algorithm has two stages, in the first stage, an elected coordinator collects information from the overlapped cells to build a channel access map that should be followed by all the overlapped cells in order to reuse the available channels without any interference between the cells. In the second stage, based on the channels determined for each cell in the first stage, each cell establishes a channel access map for its CPEs. Note that the same algorithm is used in both stages to build the channel access maps. The idea behind the used algorithm is assigning channels to the overlapped users based on a proposed function that has a logarithmic growth with the number of assigned channels for each user. This used function was selected in order to satisfy the required fairness between overlapped users. In this paper, we consider spectrum sharing among overlapped cells as well as among the CPEs in one cell. Therefore, we propose Resource Sharing Protocol for Inter/Intra WRAN communications (RSP-WRAN). This protocol is discussed in the next section.

3. Proposed protocol: RSP-WRAN

According to the IEEE 802.22 standard [15,45], time is divided into 160 msec periodic intervals, called superframes (see Fig. 4). Each superframe is further divided into 16 frames. During each frame, the BS manages the up-stream (i.e., from CPEs to BS) and down-stream (i.e. from BS to CPEs) transmissions. At the beginning of each superframe (i.e., during Superframe Control Header (SCH)), the BS communicates with its CPEs on the operating channel. So all the CPEs will be synchronized with their BS which will control their transmissions.

Fig. 4.

Superframe structure in IEEE 802.22.

Our proposed protocol RSP-WRAN is a time slotted protocol and it is based on the same superframe structure defined by the IEEE 802.22 standard. Since the maximum allowable time of occupying an available channel by a secondary user is 2 seconds [14], the channels assignment process is repeated by RSP-WRAN every 12 superframes (i.e., 1.92 seconds). In a nutshell, every 2 seconds, a new sensing process is performed to determine the available channels and the RSP-WRAN is run to allocate these channels to the CPEs according to their positions and requests. The positions of CPEs are needed to achieve frequency reuse without interference (i.e., to avoid assigning the same frequencies to neighboring CPEs). The request R of a CPE is the number of superframes (i.e., $0 ⩽ R ⩽ 12$ ) requested by the CPE to satisfy its transmission of amount of data D (in Mbits). The request R is calculated as follows. $\begin{matrix} (1) & R = \frac{D}{W \times η \times T_{s}} \end{matrix}$ where W is the bandwidth of the TV spectrum (i.e., 6 MHz), $T_{s}$ is the superframe duration (i.e., 160 msec), and η is the information rate that can be transmitted over the given bandwidth. We use $η = 1.01$ as in [58].

The RSP-WRAN protocol is explained in the following steps:

Step 1: In each beacon group BG (see Fig. 2), one BS is elected as a coordinator using the algorithm proposed in [64].

This algorithm is chosen due to its scalability (i.e., it can be efficiently employed for large groups) as well as its reliability (i.e., it has a solution for coordinator crash).

Step 2: Each CPE transmits to its corresponding BS its geographic position, request, and the sensing results (i.e., available channels).

Step 3: Each BS collects the information sent by its CPEs (in step 2) and then forwards such information to the corresponding coordinator (elected in step 1). For the BS-to-BS communications, IEEE 802.22 standard defines a protocol called Coexistence Beacon Protocol (CBP) [68]. CBP is based on transmitting beacons during a certain duration in the WRAN frame (see Fig. 4) to achieve self-coexistence among coexisting WRAN cells.

Step 4: Each coordinator in each BG collects the forwarded information from its BSs (in step 3) and uses such information to build up two tables: the overlay table, that depicts the interference map of the overlapped CPEs in the BG, and the channel table that shows the available channels for each CPE in the BG. The overlay table is an $N \times N$ matrix, where N is the total number of CPEs in the BG that require channels for transmission. The $(i, j)$ element of the overlay table is set to 1 if ${CPE}_{i}$ and ${CPE}_{j}$ are in the transmission range of each other. The channel table is an $N \times M$ matrix, where N is the total number of CPEs in the BG and M is the total number of available channels to the BG. The $(i, j)$ element of the channel table is set to 1 if ${channel}_{j}$ is available for ${CPE}_{i}$ .

Step 5: The coordinator runs the proposed spectrum sharing algorithm. This algorithm allows CPEs in the BG to use the available channels efficiently, fairly, and without interference among them. Moreover, our proposed algorithm increases the spectrum capacity by allowing channel reuse among the non-overlapped CPEs. It should be mentioned that, to consider the aggregate interference from the CPEs that use the same channel, the coordinator specifies the transmitted power for such CPEs such that a certain SINR threshold is satisfied at each CPE. In particular, the proposed sharing algorithm generates an access map for 2 seconds duration (i.e., 12 superframes). This access map determines the WRANs that are allowed to transmit on each available channel during each superframe.

The proposed algorithm is depicted in Fig. 5. For the sake of illustration, the details of this algorithm is explained via the example shown in Fig. 1. To generalize the function of the proposed algorithm, in the considered example and in Section 4, we will consider spectrum sharing among “WRAN users” that could be CPEs or BSs. In the considered example, there are six overlapped WRAN users (i.e., BSs) and the overlay table for this scenario is shown in Table 1. For the channel table of this example (see Table 2), assume that there are three available channels: Channel A, Channel B, and Channel C. Let Channel A be only available for WRAN1, WRAN4 and WRAN6, Channel B is only available for WRAN2, WRAN3 and WRAN5, while Channel C is available for all WRANs.

Fig. 5.

Flow-chart shows the proposed resource sharing algorithm.

In our example, to explain the channel assignment process during one superframe, let the number of superframes requested by the six WRANs be $[3, 3, 2, 2, 1, 1]$ . As the first step, the transmissions on channel A are scheduled as depicted in Fig. 6(a). As shown in the first column in this figure, channel A is only available for WRAN1, WRAN4, and WRAN6. So, one of these WRANs will use channel A. To determine this WRAN, the process is started by defining an assigned superframes vector AS_V to indicate the number of assigned superframes to each WRAN during the assignment process. This vector is initially set to zeros $[0, 0, 0, 0, 0, 0]$ (as shown in Fig. 6(a) and Fig. 5). As shown in the first column in Fig. 6(a), there are three possible assignments for channel A. Accordingly, as shown in the second column, there are three possibilities for AS_V. As a second step in the assignment process, a fairness parameter X is calculated for each WRAN in each possible assignment as shown in the third column of Fig. 6(a). The parameter X for ${WRAN}_{i}$ is defined as the ratio of the number of assigned superframes to the number of requested superframes for ${WRAN}_{i}$ . Since the WRANs require different number of superframes, the parameter X is considered as a good satisfaction metric. This parameter (for multiple WRANs) is used to calculate the Jain’s fairness index (FI) as shown in the last column of Fig. 6(a). The fairness index for N WRAN users is calculated as follows: $\begin{matrix} (2) & FI = \frac{| \sum_{i = 1}^{N} x_{i} |^{2}}{N \sum_{i = 1}^{N} x_{i}^{2}} \end{matrix}$ Where $x_{i}$ is the parameter X for WRAN user i.

Finally, the assignment that results in the highest FI is selected (in the case of equality, any one is selected). Accordingly, Channel A is assigned for WRAN1 during the first superframe and AS_V becomes $[1, 0, 0, 0, 0, 0]$ . After that the assignment process for Channel B is started following the same procedure used for Channel A. The step by step algorithm implementation for the Channel B and Channel C is illustrated in Fig. 6(b) and Fig. 6(c), respectively. Therefore, for our example, the final result of the assignment process during the first superframe is shown in Fig. 7. For the next superframe, the same assignment process is restarted using as input the last AS_V. The whole process is depicted by the flow chart shown in Fig. 5. As explained before, the assignment process is done for 12 superframes (i.e., 2 seconds).

Table 1

Overlay table for the scenario used in Fig. 1

	W1	W2	W3	W4	W5	W6
W1	1	0	0	1	0	1
W2	0	1	1	1	1	1
W3	0	1	1	0	1	0
W4	1	1	0	1	0	1
W5	0	1	1	0	1	0
W6	1	1	0	1	0	1

Table 2

Channel table for the scenario used in Fig. 1

	Channel A	Channel B	Channel C
W1	1	0	1
W2	0	1	1
W3	0	1	1
W4	1	0	1
W5	0	1	1
W6	1	0	1

Fig. 6.

Illustration of channel assignment process during one superframe for the network of the example considered in Section 3.

Fig. 7.

The channels assignment during the first superframe for the scenario shown in Fig. 1.

Step 6: After the assignment process for 12 superframes, the coordinator broadcasts the access map for all WRANs in the BG. The AS_V is reset to zeros $[0, 0, 0, 0, 0, 0]$ , and the assignment process restarts with a new sensing phase.

4. Performance evaluation

4.1. Simulation setup

In this section, we study the performance of the proposed protocol RSP-WRAN and compare it with the protocols proposed in [44,59], and [35]. As discussed before in Section 2, in [59], the authors proposed an inter/intra resource sharing algorithm, called 2I-RSA. This algorithm is based on a proposed function that has a logarithmic growth with the number of assigned channels for each user. According to this algorithm, the channel assignment that results in the highest value of the proposed function is selected. In [44], the authors proposed an On Demand Spectrum Contention (ODSC) protocol. In this protocol, WRAN BSs run a contention mechanism in a distributed way to access the available channels.

In [35], the Max-Min mechanism is used as a resource sharing method. The first step in Max-Min mechanism is computing the chromatic number which is the minimum number of different channels required in the network for the simultaneous transmissions without causing interference among them. The second step in Max-Min mechanism is finding the groups of WRANs that can use the same channel without interference. Finally, the available spectrum is divided into a number of bands equal to the chromatic number. Each band is assigned to each group. In the case that a WRAN group requires less than its band, the excess is shared using the same criterion by the remaining groups.

In the evaluation process, we mainly focus on two performance metrics: network throughput and fairness. To evaluate fairness, we calculate the Jain’s fairness index [48] which is computed using (2) in Section 3. These metrics are evaluated for different scenarios.

First, we examine the same scenario considered in [59]. This scenario is shown in Fig. 1. In such scenario, there are six overlapped WRANs. The used simulation time is 1.92 seconds (i.e., 12 superframes). We assume that there are two channels and they are available for all WRANs. Also, we assume that the available bandwidth is 6 MHz (i.e., one TV channel). Furthermore, we assume that the six WRANs need to send the following amount of data (in Mbit) $[6.7, 6.7, 9.6, 9.6, 11.5, 11.5]$ . Therefore, utilizing (1), the corresponding number of superframes required by each WRAN is $[7, 7, 10, 10, 12, 12]$ superframes.

The simulation results for the scenario are depicted in Fig. 8. These results show that RSP-WRAN and 2I-RSA almost have the same performance in terms of overall network throughput (see Fig. 8(c)) and allowing each WRAN to nearly send its required data (see Fig. 8(a)). However, Fig. 8(b) shows that RSP-WRAN achieves higher fairness index compared with all other protocols. This is because, as explained in Section 3, RSP-WRAN is trying to maximize the fairness index in each step of the assignment process. Also, these results show that ODSC has the worst performance. This is because ODSC is a distributed protocol; resulting in high collision and interference probability. Also, ODSC does not consider the requests of WRANs in the sharing process since it is based on a contention mechanism. On other hand, Max-Min, as RSP-WRAN and 2I-RSA, are centralized protocols. However, Max-Min does not consider a WRAN request with respect to the total requests of all WRANs in the BG. To illustrate, for the scenario shown in Fig. 1, the chromatic number is 3 (i.e., three WRAN groups: WRAN1–WRAN2, WRAN3–WRAN4, and WRAN5–WRAN6). The total available spectrum is 12 MHz (i.e., two TV channels). So, 4 MHz is assigned for each WRAN group. Therefore, by using (1), the amount of data that can be sent in 4 MHz is 7.75 Mbit (during 12 superframes). The first group (i.e., WRAN1 and WRAN2) need to send 6.7 Mbit, so 4 MHz is more than needed. The excessive band is divided among other groups. However, the requests of the second and third groups are not completely achieved. So, in Max-Min, the WRANs that have smaller requests are almost satisfied more than other WRANs.

Accordingly, Fig. 8(a) shows that Max-Min assigns WRAN3, WRAN4, WRAN5 and WRAN6 the same resources even though they have different demands. On the other hand, Fig. 8(a) shows that, in RSP-WRAN and 2I-RSA, the available resources are divided proportionally to the demands. In other words, the curves of the RSP-WRAN and 2I-RSA have the same trend of the request curve.

Fig. 8.

Performance of various resource sharing protocols for the network topology shown in Fig. 1 where 6 WRAN cells are considered and 2 channels are available to all the cells.

Fig. 9.

Fairness achieved by various resource sharing protocols in 8 WRAN cells (overlay table is shown in Table 3).

Fig. 10.

Fairness achieved by various resource sharing protocols for different number of WRAN users with 2 available channels.

Fig. 11.

Fairness achieved by various resource sharing protocols for different number of WRAN users with 5 available channels.

Fig. 12.

Performance of various resource sharing protocols for 10 WRAN cells, 3 available channels.

To verify the performance of RSP-WRAN, more simulations for different scenarios were conducted. Figure 9, Fig. 10, Fig. 11, and Fig. 12 show the performance of RSP-WRAN, 2I-RSA, and ODSC under different scenarios. It should be mentioned that Max-Min scheme cannot be evaluated under these scenarios because Max-Min scheme requires that all the WRANs should have the same available channels (which is not satisfied in the considered scenarios). This is a required condition for Max-Min scheme to compute the chromatic number.

Table 3

Overlay table for the scenario used in Fig. 9

	W1	W2	W3	W4	W5	W6	W7	W8
W1	1	1	1	1	1	0	1	1
W2	1	1	1	1	1	1	1	1
W3	1	1	1	1	1	1	1	1
W4	1	1	1	1	1	0	1	1
W5	1	1	1	1	1	0	1	0
W6	0	1	1	0	0	0	1	0
W7	1	1	1	1	1	1	1	1
W8	1	1	1	1	0	0	1	1

Figure 9 shows the effect of changing the number of available channels on the performance of the compared algorithms. In this figure, we consider 8 cells with the overlay table shown in Table 3 and the required transmission load for each cell is selected randomly as follows: $[33.9360, 2.9088, 31.9968, 3.8784, 19.3920, 32.9664, 0.9696, 8.7264]$ Mbits. Each result shown in this figure is the average of different experiments with different channel tables that are randomly built for each experiment. As we can see from this figure, our proposed algorithm RSP-WRAN achieves the highest performance.

In Fig. 10 and Fig. 11, the number of available channels is fixed to 2 and 5 channels, respectively. However, in these figures, the number of WRAN users is changed. Each result shown in these figures is the average of different experiments. In each experiment, overlay- and channel-tables as well as the required transmission load for each cell are randomly selected. These figures show that our proposed protocol RSP-WRAN achieves the highest fairness compared with the compared schemes.

Table 4

Overlay table for the scenario used in Fig. 12

	W1	W2	W3	W4	W5	W6	W7	W8	W9	W10
W1	1	1	1	1	1	0	1	1	1	1
W2	1	1	1	1	1	1	1	1	1	0
W3	1	1	1	1	1	1	1	1	1	0
W4	1	1	1	1	1	0	1	1	1	1
W5	1	1	1	1	1	0	1	0	1	1
W6	0	1	1	0	0	1	1	0	1	0
W7	1	1	1	1	1	1	1	1	1	0
W8	1	1	1	1	0	0	1	1	1	0
W9	1	1	1	1	1	1	1	1	1	0
W10	1	0	0	1	1	0	0	0	0	1

Table 5

Channel table for the scenario used in Fig. 12

	Channel 1	Channel 2	Channel 3
W1	0	1	1
W2	1	0	1
W3	1	1	1
W4	0	1	1
W5	0	1	0
W6	0	1	0
W7	0	1	1
W8	1	0	1
W9	0	0	1
W10	1	0	1

In Fig. 12, 10 WRAN cells with 3 available channels are considered. Moreover, the overlay- and channel-tables are randomly selected as shown in Table 4 and Table 5, respectively. From the results shown in Fig. 12, we can see that RSP-WRAN achieves the highest performance in terms of fairness and network throughput. All the above results can be contributed to the fact that RSP-WRAN is based on maximizing the Jain’s fairness in each step of the assignment process (as explained in Section 3).

5. Conclusions

In this paper, we proposed a new protocol called RSP-WRAN for channels assignment in IEEE 802.11b WRAN. RSP-WRAN is an inter/intra-network resource sharing protocol. When it assigns channels for WRAN users, it takes into account the interference between the overlapped WRANs as well as the interference between the users in the same cell. RSP-WRAN exploits the spatial diversity in the channel assignment to efficiently utilize the available channels and increase the network throughput. RSP-WRAN is mainly trying to satisfy the fairness among the WRAN users. The basic idea behind RSP-WRAN is employing the Jain’s fairness index in the channel assignment procedure. According to RSP-WRAN, after each sensing process (i.e., every 2 seconds duration), each WRAN user knows in which superframes and at what channels it can send data. RSP-WRAN searches in all possible assignments and selects the one that maximizes the Jain’s fairness index. Extensive simulations showed that RSP-WRAN, compared to several resource sharing techniques (2I-RSA, Max-Min, and ODSC) significantly achieves high network performance in terms of throughput and fairness.

References

M.F.

Al-Mistarihi,

Mohaisen,

Sharaqa,

M.M.

Shurman and

K.A.

Darabkh, Performance evaluation of multiuser diversity in multiuser two-hop cooperative multi-relay wireless networks using MRC over Rayleigh fading channels, International Journal of Communication Systems 28(1) (2015), 71–90. doi:10.1002/dac.2640.

Al-Zubi,

Hawa,

Al-Sukkar and

K.A.

Darabkh, Markov-based distributed approach for mitigating self-coexistence problem in ieee 802.22 wrans, The Computer Journal (2013), 1–11. doi:10.1093/comjnl/bxt092.

Al-Zubi and

Krunz, Interference management and rate adaptation in OFDM based UWB networks, IEEE Transactions on Mobile Computing (TMC) 9(9) (2010).

Al-Zubi and

Krunz, Overhearing-aware joint routing and rate selection in multi-hop multi-rate UWB-based WPANs, in: Proceedings of the IEEE INFOCOM Conference, 2010, pp. 2955–2963.

Al-Zubi and

Krunz, Cross-layer design for efficient resource utilization in WiMedia UWB-based WPANs, ACM Transactions on Modeling and Computer Simulation (TOMACS) – Special Issue on Modeling and Simulation of Cross-Layer Interactions in Communication Networks, 21(1) (2010).

Al-Zubi and

Krunz, Opportunistic routing in multi-rate multi-hop ad hoc networks, in: IEEE ICC Conference, 2012, pp. 238–242.

Al-Zubi,

Krunz,

Al-Sukkar,

Hawa and

K.A.

Darabkh, Packet recycling and delayed ack for improving the performance of TCP over MANETs, Wireless Personal Communications 75(1) (2014), 943–963. doi:10.1007/s11277-013-1401-8.

Al-Zubi,

Krunz and

Lopes, Resource utilization mechanism for multi-rate ultra-wide band networks, in: Proceedings of the IEEE GLOBECOM Conference, December 2010.

Al-Zubi,

M.Z.

Siam and

Krunz, Coexistence problem in IEEE 802.22 wireless regional area networks, in: Proceedings of the IEEE Global Telecommunications Conference (GLOBECOM), November 2009.

10.

R.T.

Al-Zubi,

Krunz and

Muqattash, Interference management distributed reservation protocol for OFDM-based UWB communications, in: IEEE GLOBECOM Conference, 2008, pp. 5228–5232.

11.

R.T.

Al-Zubi,

Krunz and

H.B.

Salameh, IMPORTANT: Integrating Multi-rate caPability into Opportunistic Routing in uwb-based Ad hoc NeTworks, Computer Communications 53(10) (2014), 84–94. doi:10.1016/j.comcom.2014.07.012.

12.

Chen,

Li and

Li, Auction-based inter bs spectrum sharing for ieee 802.22 wran systems, in: Proceedings of the IEEE Conference on Industrial Electronics and Applications, May 2009, pp. 1787–1791.

13.

Cognitive Wireless RAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications: Policies and Procedures for Operation in the TV Bands, IEEE P802.22/D0.1 Draft Standard for Wireless Regional Area Networks – Part 22.

14.

Cordeiro,

Challapali,

Birru and

Shankar, IEEE 802.22: The first worldwide wireless standard based on cognitive radios, in: Proceedings of the First IEEE International Symposium on New Frontiers in Dynamic Spectrum Access Networks (DySPAN), 2005, pp. 328–337. doi:10.1109/DYSPAN.2005.1542649.

15.

Cordeiro,

Challapali,

Birru and

Shankar, IEEE 802.22: An introduction to the first wireless standard based on cognitive radios, Journal of Communications 1(1) (2006), 38–47. doi:10.4304/jcm.1.1.38-47.

16.

Darabkh,

Al-Maaitah,

Jafar and

Khalifeh, Energy efficient clustering algorithm for wireless sensor networks, in: Proceedings of 2017 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET 2017), Chennai, India, March 2017.

17.

K.A.

Darabkh, Queuing analysis and simulation of wireless access and end point systems using Fano decoding, Journal of Communications 5(7) (2010), 551–561. doi:10.4304/jcm.5.7.551-561.

18.

K.A.

Darabkh, Evaluation of channel adaptive access point system with Fano decoding, International Journal of Computer Mathematics 88(5) (2011), 916–937. doi:10.1080/00207160.2010.485249.

19.

K.A.

Darabkh, Fast and upper bounded Fano decoding algorithm: Queuing analysis, Transactions on Emerging Telecommunications Technologies 28(1) (2017), 1–12. doi:10.1002/ett.2929.

20.

K.A.

Darabkh and

Abu-Jaradeh, Bounded Fano decoders over intermediate hops excluding packet retransmission, in: Proceedings of IEEE 24th International Conference on Advanced Information Networking and Applications (AINA 2010), Perth, Australia, April 2010, pp. 299–303. doi:10.1109/AINA.2010.141.

21.

K.A.

Darabkh,

B.N.

Abu-Jaradeh and

I.F.

Jafar, Incorporating automatic repeat request and thresholds with variable complexity decoding algorithms over wireless networks: Queuing analysis, IET Communications Journal 5(10) (2011), 1377–1393. doi:10.1049/iet-com.2010.0698.

22.

K.A.

Darabkh,

W.S.

Al-Rawashdeh,

R.T.

Al-Zubi and

S.H.

Alnabelsi, C-DTB-CHR: Centralized density- and threshold-based cluster head replacement protocols for wireless sensor networks, Journal of Supercomputing, in press.

23.

K.A.

Darabkh,

W.S.

Al-Rawashdeh,

Hawa,

Saifan and

A.F.

Khalifeh, A novel clustering protocol for wireless sensor networks, in: Proceedings of 2017 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET 2017), Chennai, India, March 2017.

24.

K.A.

Darabkh,

W.Y.

Albtoush and

I.F.

Jafar, Improved clustering algorithms for target tracking in wireless sensor networks, Journal of Supercomputing (2016). doi:10.1007/s11227-016-1898-1.

25.

K.A.

Darabkh and

Alsukour, Novel protocols for improving the performance of ODMRP and EODMRP over mobile ad hoc networks, International Journal of Distributed Sensor Networks 2015 (2015), Article ID 348967.

26.

K.A.

Darabkh,

A.M.

Awad and

A.F.

Khalifeh, Efficient pfd-based networking and buffering models for improving video quality over congested links, Wireless Personal Communications 79(1) (2014), 293–320. doi:10.1007/s11277-014-1857-1.

27.

K.A.

Darabkh,

A.M.

Awad and

A.F.

Khalifeh, New video discarding policies for improving UDP performance over wired/wireless networks, International Journal of Network Management 25(3) (2015), 181–202. doi:10.1002/nem.1888.

28.

K.A.

Darabkh and

Aygun, Improving UDP performance using intermediate QoD-aware hop system for wired/wireless multimedia communication systems, International Journal of Network Management 21(5) (2011), 432–454. doi:10.1002/nem.768.

29.

K.A.

Darabkh and

R.S.

Aygun, Performance evaluation of sequential decoding system for udp-based systems for wireless multimedia networks, in: International Conference on Wireless Networks (ICWN’06), June 2006, pp. 365–371.

30.

K.A.

Darabkh,

Ibeid,

I.F.

Jafar and

R.T.

Al-Zubi, A generic buffer occupancy expression for stop-and-wait hybrid automatic repeat request protocol over unstable channels, Telecommunication Systems 63(2) (2016), 205–221. doi:10.1007/s11235-015-0115-5.

31.

K.A.

Darabkh,

Jafar,

G.A.

Sukkar,

Abandah and

Al-Zubi, An improved queuing model for packet retransmission policy and variable latency decoders, IET Communications Journal 6(18) (2012), 3315–3328.

32.

K.A.

Darabkh,

Khalifeh,

Naser and

Al-Qaralleh, New arriving process for convolutional codes with adaptive behavior, in: Proceedings of IEEE/SSD’12 Multi-Conference on Systems, Signals, and Devices, Chemnitz, Germany, March 2012, pp. 1–6.

33.

K.A.

Darabkh and

W.D.

Pan, Queueing simulation for Fano decoders with finite buffer capacity, in: Proceedings of the 9th Communications and Networking Simulation Symposium Conference (CNSS’06), Huntsville, Alabama, 2006.

34.

F. 10-174, Unlicensed operations in the tv broadcast bands, Federal Communications Commission, Washington, DC, USA, 2010.

35.

R.Z.Z.

Fan, Spectrum allocation and medium access in cognitive radio wireless networks, in: European Wireless, 2009.

36.

Gao,

Cai,

Li,

Wei and

Chen, Credit token based dynamic resource renting and offering mechanism for cognitive radio wran bs spectrum sharing, in: Proceedings of the International Conference on Advanced Information Networking and Applications – Workshops, March 2008, pp. 296–300.

37.

Ghosh,

Roy and

Cavalcanti, Coexistence challenges for heterogeneous cognitive wireless networks in TV white spaces, IEEE Wireless Communications 18(4) (2011), 22–31. doi:10.1109/MWC.2011.5999761.

38.

Grandblaise and

Hu, Inter base stations adaptive on demand channel contention for IEEE 802.22 WRAN self coexistence, Technical proposal submitted to the IEEE 802.22 WG, Jan 2007.

39.

Grandblaise,

Moessner,

Vivier and

Tafazolli, Credit token based rental protocol for dynamic channel allocation, in: Proceedings of the International Conference on Cognitive Radio Oriented Wireless Networks and Communications, 2006.

40.

Grandblaise,

Moessner,

Vivier and

Tafazolli, Credit token based scheduling for inter bs spectrum sharing, in: 4th Karlsruhe Workshop on Software Radios, March 2006, pp. 22–23.

41.

Hani,

Salameh,

Jararweh and

Bousselham, Traffic-aware self-coexistence management in IEEE 802.22 wran systems, in: Proceedings of the IEEE Conference and Exhibition (GCC), Nov. 2013, pp. 507–510.

42.

Hawa,

K.A.

Darabkh,

L.D.

Khalaf and

J.S.

Rahhal, Dynamic resource allocation using load estimation in distributed cognitive radio systems, AEU-International Journal of Electronics and Communications 69(12) (2015), 1833–1846. doi:10.1016/j.aeue.2015.09.008.

43.

R.A.-Z.M.

Hawa,

K.A.

Darabkh and

Al-Sukkar, A self-learning mac protocol for energy harvesting and spectrum access in cognitive radio sensor networks, Journal of Sensors 2016 (2016), Article ID 9604526.

44.

Hu, On-demand spectrum contention for inter-cell spectrum sharing in cognitive radio networks, U.S. Patent No. 8780882 B2, July 2014.

45.

Hwang,

J.S.

Um,

Song,

Kim,

Park and

Kim, Design and verification of ieee 802.22 wran physical layer, in: International Conference on Cognitive Radio Oriented Wireless Networks and Communication, May 2008.

46.

IEEE 802.22 Working Group, tech. rep., http://www.ieee802.org/22/.

47.

IEEE Recommended Practice for Information Technology – Telecommunications and Information Exchange Between Systems Wireless Regional Area Networks (WRAN) – Specific requirements – Part 22.2: Installation and Deployment of IEEE 802.22 Systems, Institute of Electrical and Electronics Engineers, New York, USA, http://standards.ieee.org/getieee802/download/802.22.2-2012.pdf.

48.

Jain, The Art of Computer System Performance Analysis, John Wiley & Sons, New York, 1991.

49.

Jain,

Chiu and

Hawe, A quantitative measure of fairness and discrimination for resource allocation in shared computer systems, DEC Research, Technical Report, 1984.

50.

A.F.

Khalifeh,

AlQudah and

K.A.

Darabkh, Optimizing the beacon and SuperFrame orders in IEEE 802.15.4 for real-time notification in wireless sensor networks, in: Proceedings of 2017 International Conference on Wireless Communications, Signal Processing and Networking (WiSPNET 2017), Chennai, India, March 2017.

51.

Kim,

Oh,

Gerla and

Park, Cocast: Multicast mobile ad hoc networks using cognitive radio, in: Proceedings of the IEEE Conference on Military Communications (MILCOM), 2009, pp. 831–837.

52.

C.-H.

Ko and

H.-Y.

Wei, Game theoretical resource allocation for inter-bs coexistence in ieee 802.22, IEEE Transactions on Vehicular Technology 59(4) (2010), 1729–1744. doi:10.1109/TVT.2010.2043376.

53.

Lee,

C.S.

Hongand,

Oo and

Jung, Design of handover scheme for ieee 802.22 WRAN, in: Proceedings of the International Conference on Ubiquitous Information Management and Communication (ICUIMC), 2012, pp. 1–9.

54.

K.B.

Letaief and

Zhang, Cooperative communications for cognitive radio networks, Proceedings of the IEEE 97(5) (2009), 878–893. doi:10.1109/JPROC.2009.2015716.

55.

Mitola, Cognitive radio: Making software radios more personal, IEEE Personal Communications 6(4) (1999). doi:10.1109/98.788210.

56.

Mmary, Comparative study between wireless regional area network (IEEE standard 802.22) and wimax and coverage planning of a wireless regional area network using cognitive radio technology, Master Thesis, Communications Research Group, University of York, York, UK, 2011.

57.

Niyato,

Hossain and

Han, Dynamic spectrum access in ieee 802.22-based cognitive wireless networks: A game theoretic model for competitive spectrum bidding and pricing, IEEE Wireless Communications 16(2) (2009), 16–23. doi:10.1109/MWC.2009.4907555.

58.

Passiatore and

Camarda, Inter-network resource sharing in ieee 802.22 networks, Communications and Network 4 (2012), 111–121. doi:10.4236/cn.2012.42015.

59.

Passiatore and

Camarda, A 802.22b protocol for inter/intra wran communication, EAI Endorsed Transactions on Context-Aware Systems and Applications 2(3) (2015), 1–3. doi:10.4108/casa.2.3.e5.

60.

Passiatore and

Camarda, A p2p resource sharing algorithm (p2p-rsa) for 802.22b networks, in: International Conference on Context-Aware Systems and Applications (ICCASA), October 2014.

61.

Pawelczak,

Janssen and

Prasad, Performance measures of dynamic spectrum access networks, in: Proceedings of the IEEE Conference on Global Telecommunications (GLOBECOM), 2006, pp. 1–6.

62.

Rahman,

Hossain,

Sayeed and

Palash, Cognitive radio for broadband access in rural Africa and other developing countries, Int. J. Recent Technol. Eng. (IJRTE) 1 (2013), 161–163.

63.

Sengupta,

Chandramouli,

Brahma and

Chatterjee, A game theoretic framework for distributed self-coexistence among ieee 802.22 networks, in: Proceedings of the IEEE Global Telecommunications Conference, November 2008, pp. 1–6.

64.

Sharifloo,

Mirakhorli,

Esmaeili and

A.T.

Haghighat, A leader election algorithm for clustered groups, in: International Conference on Industrial and Information Systems, August 2007.

65.

Shurman,

Al-Mistarihi and

Darabkh, Merging dynamic address autoconfiguration and security key protocols in mobile ad hoc networks, in: Proceedings of 36th IEEE International Convention on Information and Communication Technology, Electronics and Microelectronics (MIPRO 2013), Opatija, Croatia, May 2013, pp. 441–445.

66.

M.M.

Shurman,

M.F.

Al-Mistarihi and

K.A.

Darabkh, Dynamic distribution of security keys and IP addresses coalition protocol for mobile ad hoc networks, Automatika – Journal for Control, Measurement, Electronics, Computing and Communications 57(4) (2016).

67.

M.Z.

Siam,

Al-Zubi and

Krunz, Proposing effective schemes for WRANs, International Journal of Computer and Electronics Research 3(5) (2014).

68.

Stevenson,

Chouinard,

Lei,

Hu,

Shell-hammer and

Caldwell, Ieee 802.22: The first cognitive radio wireless regional area network standard, Communications Magazine 1 (2009).

69.

Wendong,

Willkomm,

Abusubaih,

Gross,

Vlantis,

Gerla and

Wolisz, Cognitive radios for dynamic spectrum access-dynamic frequency hopping communities for efficient IEEE 802.22 operation, IEEE Communications Magazine 45 (2007).

70.

Zhao,

Sasaki and

Niwano, Link budget analysis for broadband services in ieee 802.22b. ieee p802.22 working group on wireless rans discussions, 2013, https://mentor.ieee.org/802.22/dcn/12/22-12-0071-00-000b-link-budget-analysis-forbroadband-services-in-ieee-802-22b.pdf.

	W1	W2	W3	W4	W5	W6
W1	1	0	0	1	0	1
W2	0	1	1	1	1	1
W3	0	1	1	0	1	0
W4	1	1	0	1	0	1
W5	0	1	1	0	1	0
W6	1	1	0	1	0	1

	W1	W2	W3	W4	W5	W6	W7	W8
W1	1	1	1	1	1	0	1	1
W2	1	1	1	1	1	1	1	1
W3	1	1	1	1	1	1	1	1
W4	1	1	1	1	1	0	1	1
W5	1	1	1	1	1	0	1	0
W6	0	1	1	0	0	0	1	0
W7	1	1	1	1	1	1	1	1
W8	1	1	1	1	0	0	1	1

	W1	W2	W3	W4	W5	W6	W7	W8	W9	W10
W1	1	1	1	1	1	0	1	1	1	1
W2	1	1	1	1	1	1	1	1	1	0
W3	1	1	1	1	1	1	1	1	1	0
W4	1	1	1	1	1	0	1	1	1	1
W5	1	1	1	1	1	0	1	0	1	1
W6	0	1	1	0	0	1	1	0	1	0
W7	1	1	1	1	1	1	1	1	1	0
W8	1	1	1	1	0	0	1	1	1	0
W9	1	1	1	1	1	1	1	1	1	0
W10	1	0	0	1	1	0	0	0	0	1

	W1	W2	W3	W4	W5	W6
W1	1	0	0	1	0	1
W2	0	1	1	1	1	1
W3	0	1	1	0	1	0
W4	1	1	0	1	0	1
W5	0	1	1	0	1	0
W6	1	1	0	1	0	1

	W1	W2	W3	W4	W5	W6	W7	W8
W1	1	1	1	1	1	0	1	1
W2	1	1	1	1	1	1	1	1
W3	1	1	1	1	1	1	1	1
W4	1	1	1	1	1	0	1	1
W5	1	1	1	1	1	0	1	0
W6	0	1	1	0	0	0	1	0
W7	1	1	1	1	1	1	1	1
W8	1	1	1	1	0	0	1	1

	W1	W2	W3	W4	W5	W6	W7	W8	W9	W10
W1	1	1	1	1	1	0	1	1	1	1
W2	1	1	1	1	1	1	1	1	1	0
W3	1	1	1	1	1	1	1	1	1	0
W4	1	1	1	1	1	0	1	1	1	1
W5	1	1	1	1	1	0	1	0	1	1
W6	0	1	1	0	0	1	1	0	1	0
W7	1	1	1	1	1	1	1	1	1	0
W8	1	1	1	1	0	0	1	1	1	0
W9	1	1	1	1	1	1	1	1	1	0
W10	1	0	0	1	1	0	0	0	0	1

	W1	W2	W3	W4	W5	W6
W1	1	0	0	1	0	1
W2	0	1	1	1	1	1
W3	0	1	1	0	1	0
W4	1	1	0	1	0	1
W5	0	1	1	0	1	0
W6	1	1	0	1	0	1

	W1	W2	W3	W4	W5	W6	W7	W8
W1	1	1	1	1	1	0	1	1
W2	1	1	1	1	1	1	1	1
W3	1	1	1	1	1	1	1	1
W4	1	1	1	1	1	0	1	1
W5	1	1	1	1	1	0	1	0
W6	0	1	1	0	0	0	1	0
W7	1	1	1	1	1	1	1	1
W8	1	1	1	1	0	0	1	1

	W1	W2	W3	W4	W5	W6	W7	W8	W9	W10
W1	1	1	1	1	1	0	1	1	1	1
W2	1	1	1	1	1	1	1	1	1	0
W3	1	1	1	1	1	1	1	1	1	0
W4	1	1	1	1	1	0	1	1	1	1
W5	1	1	1	1	1	0	1	0	1	1
W6	0	1	1	0	0	1	1	0	1	0
W7	1	1	1	1	1	1	1	1	1	0
W8	1	1	1	1	0	0	1	1	1	0
W9	1	1	1	1	1	1	1	1	1	0
W10	1	0	0	1	1	0	0	0	0	1