CDMA versus IDMA for Subscriber Cell Density

Transcription

1 CDMA versus IDMA for Subscriber Cell Density Asma Basharat IS Dept. MCS, National Imtiaz Ahmad hohar EE Dept. MCS, National Saeed Murtaza EE Dept. MCS, National Abstract Beyond third generation (B3G) and fourth generation (4G) communication systems require bandwidth efficiency and low compleity receivers to accommodate high data rate and large number of users per cell. This paper provides the recommendations for why Interleave Division Multiple Access (IDMA) stands out among all the present day multiple access systems. Even if IDMA is a special case of Code Division Multiple Access (CDMA), it simply trounces CDMA as far as the cell capacity is concerned. In IDMA, interleavers are used as the only means for user separation. Interleaver size can be varied for a set of users but the recommended interleaver size is 104bits at which it provides better efficiency and minimum processing delay. Investigations carried throughout this paper are while using 104 bit interleavers. IDMA ehibits the tendency to handle huge number of users. 1. Introduction With the availability of high bandwidth, advance hardware and emergence of sophisticated coding techniques, a stringent need for an advanced mobile communication system is required. Wireless communication technologies have been evolving at a rapid pace over the last few decades. To meet the evergrowing challenges of mobile communication, International Telecommunications Union (ITU) formulated recommendations for 4G mobile communication systems. According to these recommendations, 4G systems are epected to be eceedingly efficient and adaptive, ehibiting high bandwidth and power efficiency in addition to low transceiver compleity. They are also required to be fleible with respect to data rate (lin adaptation), data reliability (Quality of Service), and service provisioning as well as capable of operating on frequency-selective and fast-fading channels [0]. Various multiple access schemes have been studied for B3G and 4G communication systems. Most distinguished among them are orthogonal frequency division multiple access (OFDMA), multi carrier code division multiple access (MC-CDMA) and Multiple Input Multiple Output (MIMO) CDMA [1]. In the wae of the new era of communication systems, a special case of CDMA, with all the advantages of CDMA and mitigating its discrepancies has been proposed. This iterative chip by chip multiuser detection technique is referred to as Interleave Division Multiple Access (IDMA). [1]-[7]. CDMA is a spread spectrum technique in which data of each user is interleaved using same interleaving pattern for all users and then spread using separate orthogonal codes for each user, introducing redundancy without any coding gain. The ey feature of IDMA is user separation by means of unique interleaving patterns. Bandwidth epansion in IDMA can be completely devoted to FEC coding using low rate codes and fied spreading. This provides substantial coding gain and use of unequal power distribution enhances the system performance and allows it to approach near Shannon limits of channel capacity [], [9], [11] and [13].. IDMA transmitter and receiver principles /08/$ IEEE 50

2 .1. Transmitter Structure The bloc diagram of the IDMA transmitter is shown in figure 1 [1], [] and [4]. The input data sequence of each user is first encoded using a low rate (1/) convolution code. The encoded streams are then spread (1/8) using identical spreading sequences for all users. Subsequently, the data of each user is interleaved using a unique random interleaver. Greater the randomness in the interleaving pattern better would be the system efficiency [17], [18] and [19]. After interleaving, unequal power h is assigned to individual users and is called as the fading coefficient. Figure. Receiver structures of an IDMA scheme A global turbo-type iterative process is then applied as shown in Figure to process the LLRs generated by the ESE and DECs [7] and [1]. The global process starts from the ESE. At the beginning, if no apriori information is available, the means and variances of all of the transmitted chips are set to zero and one respectively. This indicates that each chip taes on the values of +1 and -1 with equal probability (assuming binary signaling over {+1, 1}). The ESE then produces coarse estimates and delivers them to the DECs. The DECs calculate the APP of each transmitted chip being +1 and 1, and update its mean and variance, which are fed bac to the ESE for the net iteration.[5][6]. Number of iterations are evaluated for convergence of the data and then fied for any further process. Figure 1. Transmitter structures of an IDMA scheme.. Receiver Structure The receiver structure consists of an elementary signal estimator (ESE), user specific deinterleavers and single-user aposteriori probability (APP) decoders (DECs). The multiple access and coding constraints are considered separately in the ESE and DECs. The outputs of the ESE and DECs are etrinsic loglielihood ratios (LLRs) about { (} [7], [8] and [11]: P = + 1 e( ) log,, (1) P 1 = = h + n( = 1,,..., J = 1 () These LLRs are further distinguished by subscripts, i.e., e ESE ( () and e DEC ( (), depending on whether they are generated by the ESE or DECs. 3. Subscriber Cell Density: CDMA vs. IDMA IDMA is a special case of CDMA. Generally looing at the bloc diagram one can only see that the sequence of spreader and interleaver is changed. CDMA is a spread spectrum technique which uses same interleaver for its group of users but with user defined orthogonal spreading code, while IDMA uses same spreading code for all users but with user defined interleaving pattern. CDMA provides capacity increase over narrowband multiple access wireless technologies. It consistently provides better capacity for voice and data communication than other commercial mobile technologies, allowing more subscribers to connect at any given time. However, the compleity and computational cost of CDMA multiuser detection increases with the increasing number of users. In CDMA the compleity of the system is directly related to the number of users (n); as n increases the spreading code length and the compleity related to MUD also increases. This results 51

3 in degradation of spectral efficiency. In a particular length of code limited number of orthogonal sequences can eist, usually equal to the number of bits in a code. Orthogonal spreading codes are used to mitigate the Multiple Access Interference (MAI) []. This limits the number of users per cell. In CDMA the number of subscribers is equal to the number of spreading codes. On the contrary, in IDMA system low rate codes are used for spreading and user separation is performed on the user defined interleaving pattern. Also the numbers of subscribers in IDMA are not dependant on the length of the spreading code; instead they are based on the size of the interleavers used [5]. Interleaving the coded messages before transmission results in energy of users to be spread out in time. Interleavers are mainly used to protect transmission against burst errors [4] and as a means to mitigate multiple access interference (MAI) and inter-symbol interference (ISI) [] and [3]. Greater the randomness in interleaving patterns, lesser is the probability of correlated patterns [18]. Interleavers in IDMA not only need to decorrelate the energy of individual bit sequences but they also have to decorrelate the energies of the bit patterns of different users. user specific interleavers allow number of users. Interleaver bloc size can be varied for a set of users but BER and security analysis of interleaver [17], [18] and [3] recommend that optimum performance (better efficiency and minimum processing delay) can be taen with interleaver bloc size of 104 bits. Theoretically, subscriber density per cell that can be catered by an IDMA system using an interleaver of length 104 should be 104. Interleaver shuffles the data of each user in a predefined manner in such a way that correlation between individual bits of a user does not eist and interleaver of each user is developed in order to avoid correlation of the bit sequence of individual users. There should be uncorrelation within each byte i.e. more than one bit of the same byte should not remain in the same byte location. This reduces the user group size by -3. To maintain correlation between different bytes, any bit of any byte should not maintain its location in any of the other bytes as well. Therefore each bit cannot be placed in corresponding 18 bit locations i.e a bit located at any location in one byte should not go to the same location in any other byte as well as shown in Figure 3. This means we lose the combination by further -18. Therefore the number of users per cell is reduced to 104 * -3 * -18 = 893. Figure 3. Bit Locations that can be used in interleaving To optimize the transmitted signal and maimize the average E b /N o, the interleaved data of each user is multiplied with a channel coefficient h. The channel coefficient includes both the power control factor and the channel loss. According to Shannon s Theorem of channel capacity [9], optimal channel capacity can be obtained when users are allocated unequal transmission power. To achieve minimum BER and maimum channel capacity, IDMA allocates specific power to each user. IDMA has an edge over other communication systems in that the space available for power allocation, as shown above, is 893. The power requirement for each user eeps increasing as the number of users increase. Assignment of power limits the users in a system. In the IDMA system, increase in power can be taen as grouping of number of users equal to the channel coefficient. IDMA system gives optimum performance when the output of transmitter maps on the normal distribution curve [1] and [4]. ESE optimally etracts the individual user information on the basis of unequal power distribution and mapping of the user data on normal distribution curve. User data can be mapped on normal distribution curve only if the set of users is large; tending towards infinity. As the number of users increases, the weight of h factor also increases as shown in Table 1. Table 1 carries the power allocation condition for 64 users only, the assignment of unequal powers can tae the set as 1537 number of users in the system, and this limits the subscriber cell size but still an etraordinarily large amount can be supported that surpasses the total population of the world so far. Hence theoretically IDMA provides high spectral efficiency and incredible user capacity. With the advancement in technology, computer networs are evolving from Internet Protocol version 4 (IPv4) to Internet Protocol version 6 (IPv6), in which not only every individual but every device is to have its own IP address. Wireless communication systems are following the same tuition. The net generation of wireless systems has to cater for the rising communication demands and provisioning of services for increasing number of users. IDMA recons to these challenges and taes an edge over other systems. 5

4 Table 1. Power Distribution for 64 users Power allocation for 64 users Users db Factor Users db Factor Effects of Normal/Gaussian distribution in IDMA In IDMA user detection is an iterative chip-by-chip process in which apriori and aposteriori log lielihood ratios and probabilities are calculated for each chip on the basis of its mean and variance [7] [10], [17] and [18]. ) = tanh( l ESE ( Va ) = 1 ( ) = h = 1 Va ) = h = 1 ~ ), Va )/ ), )),,, ) + σ, (3a) (3b) (4a) (4b) ) + h ) eese ( ) = h. Va ) h Va ), (5) According to (1) and denoting ( m ) ( m ) ( m ) ζ i, = ri. h i (6 ), the ith chip of the received signal can be epressed as ( m ) ( m ) ( m ) ri =. h i ζ i, ( 7 ) (m ) Applying the Central limit theorm, i, in (7) can be approimated by Gaussian random variable mean and variance. Therefore, the greater the number of users, the more i, (m ) will tend towards the Gaussian distribution [],[3]. This will improve the system performance, and reduce the bit error rate. Theoretically the IDMA system performs best and error free when the number of subscribers tend toward infinity. The fewer the users the normal distribution curve will be sewed and the performance deteriorates. On the contrary, in any other multiple access system, as the number of users increases, the system performance deteriorates and the compleity and computational cost increases. 5. Conclusion In this paper an effort has been made to carry out the analysis of the IDMA. IDMA was developed with modifications in CDMA system. Since the separation in IDMA is on the basis of interleavers, IDMA taes a lead over CDMA systems. IDMA is epected to give better performance over presently available techniques. Study shows that with the increase in the number of users [7], [9] and [10], performance of IDMA receiver improves to combat against MAI. This unique feature enables this system to approach near Shannon s limit [9]. References [1] Li Ping, Lihai Liu,. Y. Wu, and W.. Leung, Interleave-Division Multiple Access, IEEE Trans. on Wireless Communication., Vol. 5, No. 4, Apr. 006, pp [] Li Ping, L. Liu, and W.. Leung, A Simple Approach to Near Optimal Multiuser Detection: Interleave-Division Multiple-Access, Proc. IEEE WCNC 003-Wireless Communications and Networing Conf., New Orleans, LA, USA, 003, pp [3] Li Ping, Lihai Liu,. Y. Wu and W.. Leung, Interleave division multiple access (IDMA) communication systems, Proc. 3rd International Symposium on Turbo Codes & Related Topics, 003, pp [4] Li Ping, L. Liu,. Y, Wu, and W.. Leung, On interleave division multiple-access, in Proc ICC 004. Paris, France, 004. [5] Li Ping, L. Liu,. Y. Wu, and W.. Leung, Approaching the capacity of multiple access channels using interleaved low-rate codes, IEEE Communications. Letters, vol. 8, pp. 4-6, Jan, 004. [6] Li Ping, L. Liu,. Y. Wu, and W.. Leung, A unified approach to multi-user detection and space-time coding with low compleity and nearly optimal performance, in Proc. 40th Allerton Conference, Allerton House, USA, Oct. 00, pp

5 [7] L. Liu, J. Tong, and Li Ping, Analysis and Optimization of CDMA Systems with chip-level Interleavers, IEEE J. Selected. Areas of Communications, Vol.4, Jan. 006, pp [8] P. Frenger, P. Orten, and T. Ottosson, Coded-spread CDMA using maimum free distance low-rate convolutional codes, IEEE Trans. on Communications., vol. 48, pp , Jan [9] C. Berrou and A. Glavieu, Near Shannon limit error correcting coding and decoding: Turbocodes, IEEE Trans. on Communications., vol. 44, pp , Oct [10] Lihai Liu, W.. Leung, Li Ping, Simple iterative chip by chip multiuser detection for CDMA system, IEEE trans. on Communications, 003. [11] X. Wang and H. V. Poor, Iterative (Turbo) Soft Interference Cancellation and Decoding for Coded CDMA, IEEE Trans. on Communication., vol. 47, July 1999, pp [1] W.. Leung, Lihai Liu, and Li Ping Interleaving- Based Multiple Access and Iterative Chip-by-Chip Multiuser Detection, IEICE Trans. on Communications, Vol. E86-B, Dec. 003,pp [13] M. Moher, An iterative multiuser decoder for nearcapacity communications, IEEE Trans. on Communications., Vol. 46, July 1998, pp [14] Omer, I.A. hohar, Zia, Waqas, Analysis of user specific interleavers for iterative Multiuser detection system in ICIE, WEC at London. U, 007. [15] Omer, I.A. hohar, Zia, Waqas, A novel approach 4G Communication system: Multi- Carrier Interleave Division Multiple Access in 7 th IASTED International conference WOC at Montreal, Canada, 007. [16] Li Ping,. Y. Wu, Lihai Liu and W.. Leung, Interleave-Division Multiple-Access, submitted to IEEE Transactions on Wireless Communications. [17] Amina Jameel, Imtiaz hohar, A Novel Approach for Random Interleavers in IDMA system, 11 th WSEAS CSCC, Crete Island, Greece. [18] Amina ameel, Imtiaz hohar, M. N. Jafri, Ahmad Raza, Random Interleaver Design, The 5 th National Research Conference, May 4, 007. [19] ai Li, Xiaodong Wang, Guosen Yue, and Li Ping, A Low-Rate Code-Spread and Chip-Interleaved Time-Hopping UWB System,IEEE Journal On Selected Areas in Communications, Vol. 4, No. 4, April 006. [0] International Telecommunication Union, Framewor and overall obectives of the future development of IMT- 000 and systems beyond IMT-000, Rec. ITU-R M.1645, June 003. [1] S. Hara and R. Prasad, Multicarrier techniques for 4G mobile communications, Boston, NJ: Artech House, 003, The Artech House Universal personal communications series. [] M. Moher and P. Guinand, An Iterative Algorithm for Asynchronous Coded Multiuser Detection, IEEE Commun. Lett. vol., Aug. 1998, pp [3] Asma Basharat, Imtiaz hohar, Saeed Murtaza, Security Analysis of Interleave Division Multiple Access: Net Generation Competitor, ICWMC, Athens, Greece,