Analysis of Multihop DCA for Cellular Networks with Different Reuse Factors

Analysis of Multihop DCA for Cellular Networks with Different Reuse Factors

Mr. Chetan D. Jadhav*1, Prof. A. S. Joshi#2

*Sipna College of Engineering & Technology, Badnera Road, Amravati, M.S.444701, India.

#Asst. Professor, Sipna College of Engineering & Technology, Badnera Road, Amravati, M.S.444701, India.

1

Abstract

Todays cellular network utilizes dynamic channel allocation system that has threats of high call blocking probability that always yields into forced call drop or termination. To cop up with the problem, we propose an efficient dynamic channel allocation scheme, which is based on carrier to noise interference ratio. A clustered Multihop Cellular Network with FCA and Nr=7 was studied earlier. Its performance was not up to the mark. We propose here a MDCA scheme with knowledge of information about interference in the surrounding cells with different reuse factors. Simulation results shows that capacity improvement at call blocking probability is 1% for MDCA over the conventional FCA are 96% and 210% for Nr=4 and Nr=7, respectively.

Index Terms Multihop Cellular Network, Channel Assignment, Mobile Ad-Hoc Networks, Clusters.

1. Introduction

   Traditional 2G Cellular networks are expanding exponentially and have almost 4 billion of subscribers till now [1]. Essentially we have a limited resource transmission spectrum that must be shared by several users. Each cell is allocated a portion of the total frequency spectrum. As users move into a given cell, they are then permitted to utilize the channel allocated to that cell. To meet the demands of increasing wireless access Hsu and Lin pro- posed multihop cellular network (MCN) [1]. As users move into a given cell, they are then permitted to utilize the channel allocated to that cell. To incorporate flexibility of Ad-Hoc and Traditional networks MCN were proposed, which uses multihop transmission for peer to peer mobile communication [2]. Therefore, channel assignment for this promising net- work architecture becomes even more difficult.

   Recently, a number of MCN architectures have been proposed and studied. For iCAR [3], fixed ad hoc relay stations (ARSs) are deployed to enable traffic balancing. iCAR is based on the idea of diverting the traffic from congested cells

   to non-congested cells using unlicensed frequency band other than the cellular frequency band, such as the industrial, scientific and medical (ISM) band. The communication link between MSs and relay stations (RSs) uses the ISM band. In [4], Yanmaz and Tonguz found that the efficiency of iCAR is dependent on the number of available ISM relay channels (CHs). Next, UCAN [5] is to increase the user data rate and system throughput through multihop relaying using the ISM band as well. In these papers, there is no detailed description on how to select and allocate the relay CHs to a MS/RS for each hop. In [6], an ad hoc GSM (A-GSM) protocol was proposed to use the cellular frequency band for RSs to cover dead spots and increase the system capacity. However, it did not clearly address how the resources are allocated to MSs and RSs. Similarly, MCN proposed in [1] also uses the cellular frequency band to implement multihop transmission, but no detailed description of channel assignment is provided.

   In our recent work [7], clustered MCN (cMCN) is proposed and studied using fixed channel assignment (FCA). It uses cellular frequency band for traffic relaying. Results shows that cMCN with FCA can improve the channel capacity. But, FCA can not deal with problems like temporal traffic demands and hot-spot. DCA is more preferable in such varying traffic demand pattern. In addition, the proposed asymmetric FCA (AFCA) for cMCN in is limited to a reuse factor of 7. Other reuse factors are not applicable. Therefore, we propose a DCA in CMCN with Multihop transmission applicable with any reuse factor.

2. Clustered Cellular Network with Multihop Transmission

   The objective of cMCN is to achieve the characteristics of the macrocell / microcell hierarchically overlaid system by applying MANET clustering [7]. In SCNs, the BS will cover the whole macrocell with a radius of rM, while cMCN divides the macrocell area into seven microcells with a radius of rm as shown in Fig. 1. Six virtual microcells will be formed around the central microcell. We use a dedicated information

   port (DIP), located at the center of each virtual microcell, as a cluster head. The function of a DIP is to allocate CHs to the MSs within its virtual microcell, to select a MS as a RS, to determine the routing path and to help its BS on the functions of authentication, authorization, and accounting (AAA). Different from the ARSs in iCAR [3], DIPs are not involved in data relaying. Therefore, their complexity is much lower than a BS, so does the cost. Each virtual microcell can be divided into two regions: inner half and outer half. The inner half is near the central microcell.

   Figure 1. Channel assignment in cMCN with mDCA.

3. Proposed DCA with Multihop

   First, we look at the channel assignment procedure under mDCA, which is shown in Fig. 1 and described as follows.

   1. One-hop Calls: A call originated from a MS in the central microcell, e.g., M S1 in microcell A, is considered as a one- hop call. It requires one UL CH and one DL CH from the microcell A. The call is accepted if microcell A finds one free UL CH and one free DL CH. Otherwise, the call is blocked.

   2. Two-hop call: A call originated from a MS in the inner half region of a virtual microcell, e.g., M S2 in region B1 of microcell B, is considered as a two-hop call. The BS will find another MS in microcell A as a RS, e.g., RS0 in Fig. 1. For UL transmission, a two-hop call requires one UL CH from the microcell B and one UL CH from microcell A. The CHs are used for the transmission from M S2 to RS0 and from RS0 to the BS, respectively. For DL transmission, a two- hop call requires two DL CHs from microcell A; one CH is for the transmission from the BS to

      RS0 and the other is for that from RS0 to M S2. The call is accepted if all the following conditions satisfy: (i) microcell B finds one free UL CH; (ii) microcell A finds one free UL CH; and (iii) microcell A finds two free DL CHs. Otherwise, the call is blocked.

   3. Three-hop Calls: A call originated from a MS in the outer half region of a virtual microcell, e.g., M S3 in region B2 of microcell B, is considered as a three-hop call. The BS finds another two MSs as the RSs; one, RS2 , is in the region B1 and the other, RS1, is in microcell A. For UL transmission, a three- hop call requires two UL CHs from microcell B and one UL CH from microcell A. The CHs are used for the transmission from M S2 to RS2 , from RS2 to RS1 and RS1 to the BS, respectively. For DL transmission, a three-hop call requires two DL CHs from microcell A and one DL CH from microcell B, which are for the transmission from the BS to RS1, from RS1 to RS2 and from RS2 to M S3, respectively. A three- hop call is accepted if all the following conditions satisfy: (i) microcell A finds one free UL CH; (ii) microcell B finds two free UL CHs; (iii) microcell A finds two free DL CHs; and (iv) microcell B finds one free DL CH. Otherwise, it is blocked.

      Next, the channel assignment is based on the information provided in the Interference Information Tables (IITs). Two global IITs, which store, update and process the channel status, are located in the mobile switching center (MSC) for uplink (UL) and downlink (DL). We assume that every DIP is able to access to the global IITs through its coresponding BS using a separate control CH. As shown in Fig. 2, a 49- microcell network model with 7 macro cells is considered. The UL IIT contains information of the entire N UL CHs, as shown in Table I. The IIT for DL is similar and not shown here. Another table, the Interference Constraint Table (ICT), as shown in Table II, contains the information of interfering cells (including the central microcell) for each microcell. ICT is also located in the MSC.

      Different reuse factor, Nr, can be used in ICT as shown in Table II. For example, for Nr =3, the number of interfering cells is six and they are located in the first surrounding tier. Refer to Fig. 2 and Table II; the interfering cells for cell 24 are cells 17, 18, 23, 25, 30, 31. There is no special sequence for the interfering cells in Table II, which is purely based on the cell configuration in Fig. 2. By using the appropriate set of interfering cell information in ICT, we can apply mDCA with any reuse factor Nr.

      Figure 2. The simulated 49-cell network.

      The content of an IIT is described as follows. For clarity, we denote the set of interfering cells of any microcell A as I(A), which varies with Nr . The information of I(A) can be obtained from the ICT.

      TABLE – I

      Interference Information Table

      Cells	Channels
      	1	2	3		N
      0	L	L	2L		L
      1		2L	U22		U33
      2	L	L	2L		2L
      48	U22	L	U33		2L

      TABLE II

      Interference Constraints Table

      1. Used Channels: A letter U11/22/33 in the (microcell A, CH j) box indicates that CH j is a used CH in microcell A;

         U11 indicates the first-hop CH; U22 indicates the second- hop CH and U33 indicates the third-hop CH. The first-hop CH means the CH used between the BS and a MS in the central microcell. The second-hop CH means the CH used between the MS in the central microcell and the MS in the inner half region of a virtual microcell. The third-hop CH means the CH used between the MS in the inner half region of a

         virtual microcell and the MS in the outer half region of a virtual microcell.

      2. Locked Channel: A letter L in (microcell A, CH j) box indicates that one cell in the set I(A), say microcell X, is using CH j. We refer microcell X as a locking cell and microcell A as a locked cell for CH j.

      3. Free Channels: An empty (microcell A, CH

      j) box indicates that CH j is a free CH for microcell A.

      Then, the channel searching strategy is described as follows. When a new call arrives, mDCA always searches for a CH from a lower-numbered CH to a higher-numbered CH in its central microcell from the UL IIT. Once a free CH is found, it is assigned to the first-hop link. Similarly, mDCA attempts to find the UL CHs for second- or third-hop links for that call in its virtual microcell from the UL IIT if it is a multihop call. For example, for a two-hop call from M S2 in Fig. 1, mDCA finds the first-hop UL CH, j, in microcell A from Table I. Based on the ICT in Table II, CH j cannot be reused in microcell

      B. Then, mDCA searches for the second-hop UL CH (not CH j) in microcell B from Table I. The searching procedure is similar for the DL.

      Finally, the channel status for a CH j in the UL (DL) IIT is updated when it is assigned to a new call or released upon a completed call. For channel assignment, after the UL (DL) CH j in microcell A is assigned to a call, the MSC will (i) insert a letter U11/22/33 with the corresponding subscript in the (A, j) entry box of the UL (DL) IIT; (ii) update the entry boxes for (I(A), j) by increasing the L value in the UL (DL) IIT. For channel releasing, after the UL (DL) CH j in microcell A is released, the MSC will (i) empty the entry box for (A, j) of the UL (DL) IIT; (ii) update the entry boxes for (I(A), j) by reducing the L value in the UL (DL) IIT.

4. Simulation Results

   The simulated network is shown in Fig. 2. Wrap-around technique is used to avoid the edge effect. We consider a total of N=140 system CHs and simply use Nr =4 and Nr =7 to illustrate the performance of mDCA. Uniform traffic is assumed and calls arrive according to a Poisson process with a call arrival rate per macrocell area uniformly. Call durations are exponentially distributed with a mean of 1/µ. The offered traffic to a macrocell is given by = /µ. Each simulation runs until 100 million calls are processed. For the conventional FCA in SCNs, the results are obtained from Erlang B formula be

   assigning each microcell N/Nr channels.

   Figure 3. Capacity for uplink and downlink for cMCN using mDCA.

   Figure 3 shows Uplink / Downlink blocking probability Pb-u / Pb-d, for the proposed scheme with Nr=4 and Nr=7. The Pb-d is lower than Pb-u for Nr=7 while Pb-d is higher than Pb-u for Nr=4.This is due to asymmetric nature of channel assignment. For 2 hop or 3 hop calls, more call are utilised in central cell for down link transmission than for Uplink. Consequently, for Nr =7, more DL CHs can be reused with minimum reuse distance and thus, Pb,D is lower than Pb,U . On the other hand, for two-hop or three-hop calls, more CHs are used in the virtual microcell for UL transmission than DL transmission. For Nr =4, those UL CHs used in virtual microcells are better packed than DL CHs used in central microcells. Thus more UL channels can be used with minimum reuse distance, and Pb-u is lower than Pb,D with Nr=4.

   Figure 4. Capacity comparison at N=140.

   Figure 4 shows Average Blocking Probability for MDCA, and the conventional FCA in SCN for Nr =4 and Nr =7. The capacity improvements at Pb =1% for mDCA over conventional FCA are 96% and 210% for Nr

   =4 and Nr =7, respectively. Furthermore, with Nr

   =7, mDCA also outperforms AFCA with a capacity improvement factor of 22%. In conclusion, our proposed mDCA can be applied for any reuse factor by adapting the ICT and it outperforms the conventional FCA and the AFCA for cMCN substantially.

5. Conclusion

In cellular mobile communication system channel demand has both spatial and temporal variation. It is seen that during busy office hours (temporal) or in some part of the city (spatial) the channel allocation changes very frequently and rapidly. Thus an efficient channel allocation algorithm is very complex and complicated task. Dynamically changing the channels allocated to different cells enable the system to adapt to temporal and spatial distribution of channel demand. The proposed technique which is based on dynamic channel allocation will show remarkable achievement in terms of i) blocking probability

ii) forced termination probability. We shall improve the technique in future that will outperform the existing technique both during peak hours and in congested part of the city, reducing the message complexity and channel acquisition delay.

References

1. Y.-D. Lin and Y.-C. Hsu, Multihop cellular: a new architecture for wireless communications, in Proc. IEEE INFOCOM00, vol. 3, pp.

   1273-1282, Tel Aviv, Israel, 26-30 Mar. 2000.

2. Shengming Jiang, Xinhua Ling and Kee-Chaing Chua, Performance of Channel Carrying for Handoff in a DCACellular Network, Wireless Personal Comm. Journal, Vol 25, Number 3 / June, 2003.

3. H. Wu, C. Qiao, S. De, and O. Tonguz, Integrated cellular and ad hoc relaying systems: iCAR, IEEE J. Select. Areas Commun., vol. 19 pp. 2105-2115, Oct. 2001.

4. E. Yanmaz and O. Tonguz, Dynamic load balancing and sharing per- formance of hybrid wireless networks, IEEE

   J. Select. Areas Commun., vol. 22, pp. 862-872, June 2004.

5. H. Luo, R. Ramjee, P. Sinha, L. E. Li, and S. Lu, UCAN: a unified cellular and ad-hoc network architecture, in Proc. ACM MOBICOM03, pp. 353-367, San Diego, CA, 14-19 Sept. 2003.

6. G. N. Aggelou and R. Tafazolli, On the relaying capacity of next generation GSM cellular networks, IEEE Pers. Commun., vol. 8, pp.

   40-47, Feb. 2001.

7. X. J. Li and P. H. J. Chong, A fixed channel assignment scheme for multihop cellular network, in Proc. IEEE GLOBECOM06, vol. WLC 20-6, pp. 1-5, San Francisco, CA, 27 Nov. 2006.