Sultan F. Meko IU-ATC, Department of Electrical Engineering Indian Institute of Technology Bombay, Mumbai , India

Transcription

1 Optimal Relay Placement Schemes In OFDMA Cellular etworks Sultan F. Meko IU-ATC, Department of Electrical Engineering Indian Institute of Technology Bomay, Mumai 76, India ABSTRACT The Wireless services such as Skype and other multimedia teleconferencing require high data rate irrespective of user s location in the cellular network. However, the Quality of Service (QoS) of users degrades at the cell oundary. Improvement in capacity and increase in coverage area of cellular networks are the main enefits of Fixed Relay odes (FRs). These enefits of FRs are ased on the position of relays in the cell. Therefore, optimal placement of FRs is a key design issue. We propose new schemes optimal FR placement in cellular network. Pathloss, Signal to Interference and oise Ratio (SIR) experienced y users, and effects of shadowing have een considered. Our analysis give more emphasis on supporting users at the cell-oundary (worst case scenario). The results show that these schemes achieve higher system performance in terms of spectral efficiency and also increase the user data rate at the cell edge. Keywords - FR deployment, OFDMA, multihop, outage proaility, Sectoring. 1. ITRODUCTIO Increase in capacity, coverage and throughput are the key requirements of future cellular networks. To achieve these, one of the solutions is to increase the numer of Base Stations (BS) with each covering a small area. But, increasing the numer of BSs requires high deployment cost. Hence, a cost effective solution is needed to cover the required area while providing desired Signal to Interference plus oise Ratio (SIR) to the users so as to meet the demand of the future cellular networks. To achieve the high data rate wireless services, Orthogonal Frequency Division Multiple Access (OFDMA) is one of the most promising modulation and multiple access techniques for next generation wireless communication networks. In OFDMA, users are dynamically allocated su-carriers and time-slots so that it is possile to minimize co-channel interference from neighoring cell y using different su-carriers. Therefore, OFDMA ased multi-hop system offers efficient reuse of the scare radio spectrum. We consider an OFDMA-ased cellular system in which users arrive and depart dynamically. Each arriving user demands rate r. If the required rate can e provided, only then a user is accepted, otherwise it is locked. Depending on the SIR experienced y an arriving user, the BS computes the sucarriers that need to allocate to the user so as to provide the required rate. If the required su-channels (i.e., a group of sucarriers) are availale, then the user is admitted. ote that the SIR decreases as the distance etween the BS and the user increases. Thus, the users at the cell oundary can cause locking proaility to e high. Since the numer of admitted users is directly proportional to the revenue of the service provider, it is imperative to design solutions that allow accommodating a large numer of users. This motivates us to propose a Fixed Relay ased cellular network architecture that is well suited to improve the SIR at the cell oundary, and thus can possily increase the numer of admitted users. Relaying is not only efficient in eliminating coverage holes throughout the coverage region, ut more importantly; it can also extend the high data rate coverage range of a single BS. Therefore andwidth and cost effective high data rate coverage may e possile y augmenting the conventional cellular networks with the relaying capaility. Fig.1 Layout of FR ased Cellular system We consider a cellular system with six fixed relay nodes (FRs) that are placed symmetrically around the BS as shown in Fig. 1. Moile stations (MSs) in outer regions AR1 to AR6 can use relaying to estalish a etter path than the direct link to BS. The key design issues in such systems are the following: (1) How su-channels should e assigned for (a) direct MS to BS links, () FR to BS links, and (c) MS to FR links. Algorithm for this is referred to as the channel partitioning scheme. (2) How su-channels can e used across various cells. 11 P a g e

2 Algorithm for this is referred to as channel reuse scheme. Effective channel partitioning maximizes utilization of every channel in the system, and thus otains high spectrum efficiency in cellular systems [3]. For a cellular system enhanced with FRs, the main idea of channel partitioning is to optimally assign the resources to the MS-BS, FR-BS, and MS-FR links. Such intra-cell spectrum partitioning along with the channel reuse scheme not only grant the data rate demanded y each user, ut also manages the inter-cell interference y controlling the distance etween any pair of co-channel links. The concept of channel partitioning for FR ased cellular system has een discussed in [4], [5], [2], [8], [7], [14]. In [4], full frequency reuse scheme was proposed. The authors divided the cell in to seven parts and allocated six sets of sucarriers to FR link and the remaining to BS. The authors aimed at exploiting the multiuser diversity gain. However, sectoring the inner region can further reduce the co-channel interference. In [5], frequency reuse scheme was proposed ased on dividing the outer region in to six and also sectoring the inner region. In [2], a preconfigured relay channel selection algorithm is proposed to reuse the channels that are already used in other cells on the links etween FRs and MSs. This scheme may suffer from high cochannel interference on FR-MS links. In [7], [14], frequency partitioning and reuse schemes for cellular WLA systems with moile relay nodes are proposed. Relays are moile as the MSs themselves act as a relay for other MSs. Since the relay is moile, the channel etween relay and the BS can change, which will result in a large numer of interrelay handoffs. Furthermore, MSs acting as relays may not e cooperative ecause of the power consumption and the security issues. In [8], a coverage ased frequency partitioning scheme is proposed. The scheme assumes that the relay nodes are placed at a distance equal to the two-third of the cell radius, and does not consider optimal relay placement. Some other proposals for frequency management include the use unlicensed spectrum [12], and the use of directional antennas [1]. In our paper [6], we proposed a channel partitioning and a channel reuse scheme for increasing system capacity to support some preceding standards like Gloal System for Moile Communications (GSM). In this paper, we extend the channel partitioning and channel reuse scheme for OFDMA cellular networks which results in increased system capacity and spectral efficiency. We also consider the optimal relay placement ased on different parameters. We show that with the appropriate relay placement, the system performance can e improved significantly. As a result, the numer of users that can e accommodated in the system can e maximized while providing each user with its required rate. The paper is organized as follows. In Section 2, we descrie our system model. In Section 3, we propose our relay placement schemes ased Path-loss, SR and Shadowing. In Section 4, we evaluate the performance of the proposed schemes using numerical computations and simulations. In Section 5, we conclude the paper. 2. SYSTEM MODEL AD DESCRIPTIO 2.1 System Configuration We consider a cellular system consisting of regular hexagonal cells each of edge length D. Each cell has a BS and certain numer of FRs situated symmetrically around the BS at a distance d r from the center and dm from the cell edge as shown in Fig.1. Let the total andwidth availale for the downlink e W units. Let the minimum SR/SIR experienced y the user at furthest location from BS/FR e γ 1. Similarly, let the minimum SR/SIR experienced y FR at furthest location from BS e γ 2, i.e, oth MS and FR are placed at a location that they experienced minimum SR or SIR. Therefore, these locations are the effective oundary for each link (i.e MS-BS, FR-BS and FR-BS). For a given MS position, let the distance from BS e d * and nearest FR e d * m. If d * d, then MS communicates with the BS directly; otherwise it communicates through the nearest relay using two hop route. Because of the specified routing scheme, a cell can e partitioned into seven regions as shown in Fig.1. We define the region covered y BS as the inner region (A 1 ) and the region covered y FRs as outer region (A 2 ). Outer region is further divided into A 2k, for k1,..., 6. All the MSs in region A 1 communicate directly to the BS, and the MSs in k th A 2 region communicate to BS through relay k th FR. Both the numer of FRs (K) and the size of each region in a cell (d * m and D) are determined y the optimal relay placement algorithm. 2.2 Channel Partitioning and Reuse Scheme The channel partitioning scheme, partitions the downlink andwidth into 2K+1 orthogonal segments, viz. W 1, W 2,k and W 3,k for k 1,...,K. The and W 1 is used y the MSs in region A 1, the and W 2,k is used y k th FR to communicate with the BS, and the and W 3,k is used y the MSs in k th A 2 region to communicate with k th FR. Let W 2 6 W k1 2,k 6 W k1 3,k and W 3. Because of the channel partitioning, there is no intra-cell interference, and the system performance is mainly determined y inter-cell co-channel interference. For channel reuse scheme, we assume that the frequency reuse distance is 1, i.e., each cell uses the complete andwidth W for the downlink communication [11]. In each 12 P a g e

3 cell, inner region uses the same and W1. While k th FR uses and W 2k to communicate with BS and MS in k th A 2 region uses W 3k and to communicate with k th FR. 2.3 Propagation Model Wireless channel suffers from fading. Fading is mainly divided into two types, slow and fast. Slow fading is due to path-loss and shadowing, while fast fading is due to multi-path. In this paper, we assume that the code lengths are large enough to reveal the ergodic nature of fast fading. Hence, we do not explicitly consider multi-path effect. We focus on the path-loss and shadowing in the analysis. Because of the path-loss, the received signal power is inversely proportional to the distance etween the transmitter and the receiver. In general, the path-loss P L etween a transmitter and a receiver is given as, P L P T G P T G R 4πf R c 2 d d γ (1) where P T is the transmitted power; P R is the received power, G T and G R are the antenna gain of transmitter and receiver respectively; f is the carrier frequency, c is the speed of light; d * is separation etween the transmitter and receiver; d is the reference distance, and γ > is the path-loss exponent [11]. the BS-MS, BS-FR and FR-MS links, henceforth, we use oth terms interchangealy. On these three links, we assume that each link has minimum SR requirement (threshold). When the received SR exceeds a threshold, the message is correctly decoded. The distance of user from BS/FR, where the received SR equals the threshold is defined as the effective radius of BS/FR; it inturn determines the optimal FR location. In the next susections, we descrie our proposed methods to deploy FRs optimally ased on path-loss and SR. 3.1 Relay Placement ased on Path-loss In this susection, we propose a simplified relay placement algorithm that depends on the pathloss. Path-loss is signal attenuation etween a source (BS/FR) and a receiver (MS) which depends on the propagation distance. 3. RELAY PLACEMET SCHEMES Improvement in capacity and increase in coverage area are the main enefits of FRs. These enefits of FRs are ased on the position of relays in the cell. Deploying FRs around the edge of the cell help the edge users. However, when they are placed at inappropriate locations, may cause interference to the edge users of the neighoring cell. Therefore, optimal placement of FRs is a key design issue. Consider the downlink scenario where the BS encodes the message and transmits it in the first time slot to neary MSs and FRs. FRs transmits the message to MSs at the cell oundary in the second time slot. FRs are either Decode-and- Forward (DF) type, which fully decodes and reencodes the message, or Amplify-and-Forward (AF) type, which amplifies and forwards the Message to MSs in the second hop. ote that the reverse will e for uplink scenario. In downlink scenarios, we consider non-transparent type relays, i.e., MSs in the first hop communicate to BS while MSs in the second hop communicate only to FRs [15]. In this section, we propose two types of Relay placement schemes to deploy FRs optimally in the cell. We analytically determine the position of FRs within the cell so that the QoS requirement of each user is satisfied. In each scheme we evaluate the signal strength on the three links (BS-MS, BS-FR, and FR-MS links). The term three links denote Fig.2 Layout of FR enhanced Cellular system illustrating the coverage of MS-BS, FR-BS and MS-FR regions. We consider a simplified cellular configuration consisting of BS and FRs as shown in Fig.2. Initially, FRs are placed at random position in a cell. MS moves from BS towards the cell oundary along a straight-line trajectory. Determination of optimal position is ased on the received signal strength and distance measurements along the path. MS evaluates the received signal strength till the signal strength ecomes equal to the preset threshold value. As the MS moves away from the BS/FR, if the received signal strength decreases elow a threshold value and the MS is not ale to communicate with the BS. At this position where the received signal strength is too weak, appropriately placed FR can enhance the signal quality of the MS. ow, we descrie our path-loss ased FR placement algorithm as follows: Referring to Fig.2, let P BS and P FR e the power transmitted y BS and FR respectively. Let P e the power received y 13 P a g e

4 MS located at the distance (d) from BS; P r e the power received y FR located at the distance (d r ) from BS; P m e the power received y MS located at the distance (d m ) from one of FRs. When MS follows straight line trajectory, then BS, MS and FR are collinear and the radius of the cell D is computed as D d + 2d m and also d r d + d m. Assume that the threshold P P m and P P r, the path-loss (1) is redefined for the three links as: transmitted signal is diffracted due to a tree or the top of a uilding along the path of propagation. Let Г BM, Г BR and Г RM e the SR experienced y the user on BS-MS, BS-FR and FR-MS links respectively. In this paper the term SR is used to denote the effects of noise and shadowing, henceforth, we use SR to represent the effect of SR plus shadowing. This algorithm is ased on evaluation of SR on three links as follows, P K 1 P BS d d P r K 2 P BS d d r P m K 3 P FR d d m γ (2) γ r (3) γ m (4) where K 1, K 2 and K 3 are constants; γ, γ r and γ m are the path-loss exponent on BS-MS, BS-FR and FR-MS links respectively. Based on (2), (3), (4) and simple algeraic manipulation, the ratio P BS P FR m dγ P γ r d r γ r (5) d r γ P r d (6) m remain constant. Based on these expressions, we define the optimal radius for BS-MS link or direct link (d) and that of FR-MS link or relaying link (d m ) as, d max d> d, s. t satisfying te constraints (7) in Equ. 5 and 6 d m argmax d m (, d r d) d m, s. t satisfying te constraints (8) in Equ. 5, 6 and 7 We choose threshold greater than the minimum signal strength elow which call drops. Thus all admitted calls never drop. Since the effects noise, shadowing and interference from neighoring cells are not considered in this model, this algorithm provides the simplified estimate of FR placement scheme. In the next susection, we propose a scheme that considers the effect of noise and shadowing. 3.2 Relay Placement ased on SR and Shadowing Relay placement algorithm ased on SR and shadowing considers the case of isolated cell, where interference from neighoring cells is not considered. In addition to signal attenuation due to path-loss, we consider the effects of noise and shadowing. Shadowing effects happen when a Г BM Г BR Г RM P BS γ log d + ξ (9) P BS γ r log d r + ξ r r () P FR γ m log d m + ξ m m (11) ote that we use suscript, r and m in this paper to refer to the three links BS-MS, BS-FR and FR- MS respectively. ξ, ξ r and ξ m are Gaussian random variale with standard deviation of σ, σ r and σ m on the three links mentioned aove. Similarly,, r and m denote the thermal noise. This algorithm evaluates the SR of each link. We consider SR as a decision parameter for the success/failure of a transmission on a link; the received message is correctly decoded when the SR experienced y MS on each link exceeds a threshold SR. Hence, the effective radius of the three links can e otained from the threshold SR on the three links. Let the threshold SR on BS-MS, BS-FR and FR-MS links are denoted as Г, Г r and Г m respectively. Accordingly, the proaility of successful decoding with direct transmission over BS-MS (Pr ) is given as, P r Pr Г BM > Г Pr P BS γ log d + ξ > Г Pr ξ > γ log d + + Г P BS (12) Q γ log d + + Г P BS σ We consider that for a given user position, if proaility that the received SR aove the threshold is 95%, then the user has good link with the BS. Therefore such users do not require relaying support. The effective radius for direct link (d) is the maximum distance where the aove criteria is satisfied. Hence, optimum value of effective radius for direct link (d) is given as, d max d> d, max γ log d + +Г P BS σ (13) s. t d > ; P r.95, Similarly, the proaility of successful decoding on BS-FR link ( P rr ) and on FR-MS link ( P rm ) can e expressed as, 14 P a g e

5 P rr Q γ r log d + r +Г r P BS σ r (14) P rm Q γ m log d + m + Г m P FR (15) σ m The equivalent SR over two-hop transmission depends on the type of relaying scheme. In this paper we consider Decode and Forward Relaying (DF) and Amplify and Forward Relaying (AF) relaying schemes. DF s. t P suc.95 ; P r.95. (19) Amplify and Forward Relaying In amplify and forward relaying scheme, the AF relay amplifies the analog signal received from the BS and transmits an amplified version of it to the MSs [14]. Let the experienced SIR for this scheme e Г AF, then it is computed as follows Decode and Forward Relaying Consider a downlink scenario, where each FR decodes the signal received from BS-FR link and re-transmits to MS on the FR-MS link. In this scheme, the end to end rate achieved from BS to MS is determined y the minimum rate achieved among the rates of BS-FR and FR-MS links, i.e., if the required rate for this scheme is R DF, then 2R DF min log Г BR, log Г RM (16) If the required rate is not achieved on either of the link, then user is said to e in outage. Let the outage proaility in DF scheme e Г DF out, it can e expressed as P DF out Pr min log Г BR 2R 1 Pr min log Г BR 2R 1 P suc, log Г RM, log Г RM DF (17) If user does not go into outage, this means there is successful transmission on BS-FR and FR-MS DF links. Let the P suc e the proaility of successful transmission on DF scheme, then it can e computed as [9], P DF suc Pr Г BR > Г r. Pr Г RM > Г m Q γ m log d m + m + Г m P FR σ m. Q γ r log d m + r + Г r P BS σ r P rr P rm (18) We consider the criterion P DF suc 95% for deploying FRs in cellular networks. Therefore, the effective radius (d m ) for relaying link (FR-MS) will e optimal distance which 5% outage proaility is satisfied, i.e., P DF out.5. Hence, the optimum radius for FR-MS link is given as, d m argmax d m (, d r d) argmax d m, d r d d m, P FR m Г m σ m.q ( P r m ) γ m Г AF Г BR ГRM Г BR + ГRM +1 (2) Let Г e the threshold SIR. If the experienced SIR falls elow the threshold, the user is said to e in out outage and outage proaility is given as P AF out for AF scheme [14]. P out AF Pr Г BR Г RM Г BR + Г RM + 1 Г (21) It is difficult to otain the exact closed-form solution for outage proaility of (21) []. However, several literatures give the closed-form lower ound and upper ound. We are interested in upper ound solution to compute the effective radius of relaying link. Accordingly, the optimum radius for relaying link ( d m ) due to AF relay is given as: d m argmax d m, d r d argmax d m (, d r d) s. t d m, P FR m Г m σ m.q ( P r m ) γ m P AF suc.95, (22) 3.3 Optimal umer of FRs Assume that users are distriuted uniformly in the cell. Hence, the approximate numer of FRs required in a cell can e otained y computing the area of inner region A 1 and that of outer region A 2. After simple mathematical steps, the numer of FR required in a cell is also computed as FR 4( d d m + 1). Since the entire cell radius is defined as D d + 2dm, the numer of FR that can support the users in outer region depend on optimal value of d and d m. We choose optimal value of d and d m that minimizes the numer of FR used in the cell. The optimization equation is descried as, FR arg min FR, s. t d, d m and d m satisfying Equ. 19 and 22, d and d m satisfying Equ. 8 (23) 3.4 Outage performance analysis Co-channel interference and shadowing effects are among the major factors that limit the 15 P a g e

6 capacity and link quality of a wireless communications system. In this section, we use Gauss-Markov model to evaluate the statistical characteristics of SIR in Multihop communication channels. By modeling SIR as log-normally distriuted random variale, we investigate the performance of relay placement schemes discussed in Susection 3.1 and 3.2. We make a comparison etween the aove models in terms of performance evaluation where outage proaility is a QoS parameter. Further more the result is used to find the more realistic way of channel partitioning and relay FR placement schemes. The co-channel uplink interference to the BS/FR of interest is assumed to e from MSs or FRs that links to first tier or upper tier cells (MS i s or FR K s ). Including the shadowing effect on the three links, the SIR on each link can e descried as; Г BM Г BR Г RM ξ d γ ξr d r γr ξm d m γm d d i d d ri 1 d γ i ξi γ ξi ξ d d mi 1 d γr ri ξri γr ξri ξr 1 d γr mi ξmi γm ξmi ξm (24) (25) (26) where d and d r are the location of desired MS and FR from the BS on direct link, d m is the location of desired MS from desired FR under the second hop. Similarly, d i and d ri are the location of co-channel interferer MS and FR from the BS, while d mi denotes the location of co-channel interferer MSs from FR that is associated to BS. Shadowing for the desired links are denoted as ξ, ξ r and ξ m For interfering links shadowing is expressed as ξ i, ξ ri and ξ mi to denote the interfering link of MS-BS, FR-BS and MS-FR; in all cases, i {1,..., 18}. Basically, the outage proaility analysis for the three links (MS-BS, FR-BS and MS-FRs) is similar to the expression given in Section 3.2. However, interference from neighoring cells is considered here. We compute the mean and standard deviation of oth desired and interference signal ased on Fenton-Wilkinson s and Schwartz-Yeh s method [13]. Shadowing is usually represented y i.i.d. log-normal model in wireless multi-hop models. However, shadowing paths are correlated. Hence, we consider correlation among interferers and also the correlation that may exist etween interferer and desired signals 4. RESULTS AD DISCUSSIO In this section, we present the analytical and simulation results to illustrate the performance of our proposed FR placement algorithms. We use Matla simulation for modeling cellular networks under varying channel conditions. We analyze the performance of relay placement schemes ased on path-loss and SR as descried in previous section. The list of simulation parameters are mentioned in Tale 1. Tale 1. List of the simulation parameters. Parameters Values Carrier Frequency 5GHz System Bandwidth (W) 25.6MHz Standard Deviation 8, 5, 8 db ( σd, σrand σm) Correlation Coefficient.5 Path-loss Exponent 3.5, 2.5, 3.5 (γd, γr and γm) BS Transmit Power (P BS ) dbm FR Transmit Power (P FR ) 2dBm MS transmit Power (P MS ) 2dBm Threshold (Г) -dbm Thermal oise () -dbm The optimal FRs placement results are summarized in the Tale 2. Tale 2. Results of optimal FRs placement ased on path-loss and SR. Parameters Based on Path-loss Based on SR d r 212m 276m d 1611m 1522m d m 9m 554m FR 6 6 Fig. 3 illustrates the outage proaility of downlink cellular network where the optimal FR placement scheme is ased on path-loss. This figure compares the outage proaility of relay enhanced cellular system of direct link and relay link along with the scheme without FRs. Furthermore, Fig. 3 demonstrates that outage proaility is significantly improved in relay enhanced cellular system with optimal FR placement. As it can e shown in this figure, at SIR threshold of db, the outage proaility of the scheme without FRs is %. But, the outage proaility of our proposed scheme for FR-MS link is nearly %. This means that due to optimal placement of FRs, outage proaility at cellular oundary improves y %. Since the users at the cell oundaries dominate the system 16 P a g e

7 Outage Proaility (%) Outage Proaility (%) Outage Proaility (%) Outage Proaility (%) ased on Pathloss Sultan F. Meko / International Journal of Engineering Research and Applications performance, our optimal relay placement scheme significantly improves the outage proaility of MSs of second hop link. 3 2 Simulation of cellular radio system Sectorization Fig. 3. Outage proaility versus SIR threshold of DF relay enhanced cellular system with no sector in the inner region; optimal FR placement is ased on path-loss. 3 2 without FR With FR (direct link MS-FR) With FR (2nd hop link MS-FR) Simulation of cellular radio system Sectoring (DF) FR placement at 2/3 of total radius Optimal FR placement ased on pathloss Optimal FR placement ased on SR Fig. 5. Outage proaility versus SIR threshold of DF relay enhanced cellular system with 12 sectoring in the inner region; proposed optimal FR placement schemes are compared with FRs location at 2/3 rd of cell radius. Fig. 4 illustrates the outage proaility versus SIR threshold for DF relay where the inner region of the cell is sectored in to sectoring (refer Fig.1). We compare our proposed optimal FR placement schemes with a scheme that places FRs at 2/3 rd position of the cell radius. The results show that our proposed schemes achieve significant improvement on the performance of the cellular network. In Fig.4, at SIR threshold of -2 db, the outage proaility of a system when FRs are placed at 2/3 rd of cellular radius is % whereas, the outage proaility of our proposed scheme for FR-MS link is only 2%. However, there is no significant difference in performance among our proposed schemes. As can e shown in Fig.4, Fig.5 and Fig.6, comparing the performance of our two proposed schemes, the outage proaility of optimal FRs placement scheme which depends on path-loss and that of the scheme which depends on SR are nearly equal. Hence, either of the two proposed schemes can e implemented for optimal FRs placement to achieve the same QOS requirement. Simulation of cellular radio system Sectorization (DF) Relay placement ased on pathloss Relay placement ased on SR Relay placement at 2/3 of total radius Fig. 4. Outage proaility versus SIR threshold of DF relay enhanced cellular system with sectoring in the inner region; proposed optimal FR placement schemes are compared with FRs location at 2/3 rd of cell radius. 3 2 Simulation of cellular radio system 12 Sectoring Optimal Relay placement ased on pathloss Optimal Relay placement ased on SR Relay placement at 2/3 of cell radiusased on pathloss Fig. 6. Outage proaility versus SIR threshold of DF relay enhanced cellular system with no sectoring in the inner region; proposed optimal FR placement schemes are compared with FRs location at 2/3 rd of cell radius. Fig.4, Fig.5 and Fig.6 also illustrate that sectoring the BS-MS direct link of the cell significantly improves the outage performance of the cell in which FRs are placed at 2/3rd of the cell radius while sectoring has no significant effect on the cellular networks that use the proposed optimally FRs placement schemes. Even though sectoring lowers the effect of co-channel interference, it also degrades system capacity. 17 P a g e

8 Outage Proaility (%) ased on Model SR Outage Proaility (%) ased on Pathloss Sultan F. Meko / International Journal of Engineering Research and Applications Fig.7 shows the performance of path-loss ased optimal FRs placement schemes for 6 and 18 co-channel interferers. We evaluate the performance of the two hop cellular network with AF and DF relay. Since AF relay scheme amplifies the required signal and noise, the cellular network performance decreases with the increase in numer of co-channel cells. We also oserve that, outage proaility of the system at -2 db SIR threshold is improved y % when the co-channel decreases from two tire (18 cell) to one tier (6 cells) cell. However such effect is not significantly oserved in DF relaying scheme Fig.7. Outage proaility versus SIR threshold of AF and DF relay enhanced cellular system with one tier and two tier co-channel cells; Optimal FRs placement is ased on path-loss 3 2 Simulation of cellular radio system Sectorization DF Relay 6 cochannel cells AF Relay 6 cochannel cells Simulation of cellular radio system (AF Relay) DF Relay 18 cochannel cells AF Relay 18 cochannel cells no sectoring 12 sectoring sectoring Fig.8. Outage proaility versus SIR threshold of AF relay enhanced cellular system with, and 12 sectors in the inner region of the cell; optimal FRs placement is ased on SR. Fig.8 shows the effect of sectoring on AF relay in which the optimal FRs placement is ased on SR. Similar to aove results, sectoring can improve the outage proaility of users y lowering the effects of co-channel interference, ut decreases system capacity. 5. COCLUSIO This research work proposed two optimal FRs placement schemes which are ased on pathloss and SR. The proposed schemes also compute optimal numer of FRs that are required to enhance the QoS of cellular system. Our schemes also investigate effects of sectoring the inner region of the cell on optimal FR placements. Sectoring the inner region for improves the outage proaility. Our optimal relay node placement schemes significantly reduce the outage of the cellular system and provide etter QoS for users at cell oundaries. 6. ACKOWLEDGMET This research work is supported y India-UK Advanced Technology Centre (IU-ATC) of Excellence in ext Generation etworks Systems and Services. References [1] Zaher Dawy, Sami Arayssi, Irahim Adel ai, and Ahmad Husseini. Fixed relaying with advanced antennas for CDMA cellular networks. In IEEE GLOBECOM proceedings, 26. [2] H. Hu, H. Yanikomeroglu, and et al. Range extension without capacity penalty in cellular networks with digital fixed relays. IEEE Gloecom, Dec 24. [3] I. Katzela and M. aghshineh. Channel assignment scheme for cellular moile telecommunication systems: A comprehensive survey. IEEE Personal Communication, pages 31, June [4] Jian Liang, Hui Yin, Haokai Chen, Zhongnian Li, and Shouyin Liu. A novel dynamic full frequency reuse scheme in OFDMA cellular relay networks. In Vehicular Technology Conference (VTC Fall), 211 IEEE, pages 1 5, sept [5] Min Liang, Fang Liu, Zhe Chen, Ya Feng Wang, and Da Cheng Yang. A novel frequency reuse scheme for ofdma ased relay enhanced cellular networks. In Vehicular Technology Conference, 29. VTC Spring 29. IEEE 69th, pages 1 5, april 29. [6] Sultan F. Meko and Prasanna Chaporkar. Channel Partitioning and Relay Placement in Multi-hop Cellular etworks. In Proc. IEEE ISWCS, 7- Sept. 29. [7] S. Mengesha, H. Karl, and A.Wolisz. Capacity increase of multi-hop cellular WLAs exploiting data rate adaptation and frequency recycling. In MedHocet, June P a g e

9 [8] M. Rong et al P. Li. Reuse partitioning ased frequency planning for relay enhanced cellular system with LOS BS-Relay links. IEEE, 26. [9] A. Papoulis. Proaility, Random Variales, and Stochastic Processes. 3rd edition, McGraw-Hill,1991 [] V. Sreng, H. Yanikomeroglu, and D. Falconer. Coverage enhancement through two-hop relaying in cellular radio systems. IEEE WCC, 2: , Mar 22. [11] D. Tse and P. Viswanath. Fundamentals of Wireless Communications. Camridge University Press, 25. [12] D. Walsh. Two-hop relaying in CDMA networks using unlicensened ands. Master s thesis, Carleton Univ., Jan 24. [13] Jingxian Wu,.B. Mehta, and Jin Zhang. Flexile lognormal sum approximation method. In IEEE GLOBECOM, pages , Dec 25. [14] H. Yanikomeroglu. Fixed and moile relaying technologies for cellular networks. In Second Workshop on Applications and Services in Wireless etworks, July 22. [15] Sultan F. Meko. Impact of Channel Partitioning and Relay Placement on Resource Allocation in OFDMA Cellular etworks, International Journal of Wireless & Moile etworks (IJWM) Vol. 4, o. 3, June P a g e