4.1. System Model and Transmission Schemes
Figure 4.1: Block diagram of the whole SC-FDMA system
Figure 4.2: The subcarrier occupations between different users.
41
The figure 4.1 shows that all users transmit their data over the same frequency resource simultaneously, but the key that those signals from different users will not interfere each other because they map the frequency domain symbols to the allocated subcarriers and transmit a null value on the subcarriers that are not allocated to it, leaving them to be used by other users.
(as shown in Figure 4.2)
4.2. Problem Formulation
Assumption
1. Each user is matched to exactly one resource block, which consist of 1 or more subcarriers.
2. All users require the same rate and have the same power requirement.
The aim of dynamic resource block allocation is to give the user channel resource such that the performance of the whole system (as overall BER is concerned) will be optimized.
First we define the gain matrix C is given by
42
And our goal is to maximize the utility function below to achieve the best performance for the whole system.
Algorithm Greedy method
1. Define 𝐾 as the set of unassigned resource blocks.
2. Start from the first user, search the gain in gain matrix C to each resource block in 𝐾.
Find the resource block with the highest gain and assign the resource block to this user.
3. Remove the resource block from 𝐾.
4. Move on to the next user and perform steps 2 to 3 again, until all the users have been allocated a resource block.
Figure 4.3: A simple example of greedy method
In Figure 4.3, the user from the first to the last searches the highest gain from the corresponding row of the gain matrix to allocate the resource block.
43
4.4. Maximum Greedy Method
Algorithm Maximum greedy method
1. First define different solution spaces (the user priority of applying the Greedy Method).
2. For each unique solution space, perform the greedy algorithm in Section 3.3 and save the solution.
3. Find the best one from saved greedy solutions.
Figure 4.4: A simple example of maximum greedy method
In Figure 4.4, there is an example that allocate the resource block to the user from the last to the first (another solution space) and searches the highest gain from the corresponding row of the gain matrix to allocate the resource block.
44
4.5. Proposed BER-Enhanced Greedy Method
Algorithm Maximum greedy method
1. Define 𝑼 as the set of all unallocated users and 𝑲 as the set of unassigned RBs.
2. Calculate the total channel gain of user 𝑢 for all users in 𝑼.
𝑚𝑢 = ‖𝐾‖1 ∑ 𝑐𝐾 𝑢𝑘, ∀𝑢 ∈ 𝑼
3. Find the user 𝑢̃ with smallest total channel gain, and then search the highest gain 𝑐𝑢̃𝑘̃
from the corresponding row of the gain matrix 𝑪 .
4. Then assign the resource block 𝑘̃ to the user 𝑢̃ and remove this RB from 𝑲.
5. Remove user 𝑢̃ from 𝑼
6. Repeat step 2 to 5 until 𝑼 ∈ ∅
Figure 4.5: A simple example of BER-enhanced greedy method
In Figure 4.5, the resource block assignment starts from the user with smallest total channel gain and assign the highest RB to the user, then followed by the second smallest until all users are allocated a RB.
45
4.6. Comparison of Computation among these three methods
Table 4.1: Comparison of computation complexity among these three methods.
(assume U=K, the number of user is the same with the number of resource block)
Algorithm # of Computation Complexity
Greedy method ∑(𝑘 − 1) =
Table 4.1 illustrates the computation complexity of these three methods. For the greedy method, it takes K-1 comparisons to find the highest gain for the first user and linearly decreasing by 1 for assigning each user a resource block. It is clearly that the total computations is ∑𝐾𝑘=1(𝑘 − 1). For the maximum greedy method, it simply a times of the greedy method. For the BER-enhanced method to find the first user to assign a RB, we need to calculate the total gain (takes K-1 computations) of K users and find the one with smallest total gain (takes K-1 computations). And then find the highest gain (takes K-1 computations) of the user to allocate a RB. So the total number of computation can be derived as ∑𝐾𝑘=1(𝑘2+ 𝑘 − 2).
46
4.7. Simulation
4.7.1. Simulation parameters
Table 4.2: Simulation parameters of Chapter 4.
System bandwidth 10M (Hz)
Subcarrier frequency 2G (Hz) Data modulation format BPSK M, Transmitter IFFT size 512
SNR 1-30 (dB)
U, Number of users 4/8/16/62/64/128/256
Equalization MMSE
Multipath fading channel PedA. channel Number of iteration 10^6
4.7.2. Simulation results
Figure 4.6: The effect on BER performance for different SNR
47
Figure 4.7: The effect on BER performance for different number of users
4.8. Conclusion of Chapter4
BER-Enhanced Greedy Method has better performance in BER because the algorithm is to find a solution space that can enhance the BER performance, but not necessarily better in total system throughput (or other performance of the whole system).
There is a trade-off between the quality of the solution and the complexity. This is very intuitive that if we spend more computation to compare and choose the better solutions.
For these three greedy methods, when the number of users increases, the BER performance will be better because the channel will be divided into smaller resource blocks.
So resource blocks can be allocated better according to the channel.
48
REFERENCES
[1] 3rd Generation Partnership Project, Technical Specification Group Radio Access Network; Physical layer aspects for evolved Universal Terrestrial Radio Access (UTRA), 3GPP Std. TR 25.814 v7.0.0.
[2] Islam M.R. , Irfan M., Ullah N., Ullah S., Fahim S.M. and Azad I.I. ,”
Non Binary SC-FDMA for 3GPP LTE uplink”, Computing, Communications and Applications Conference (ComComAp), p198-203, Oct. 2012 .
[3] de Temino L. , Berardinelli Gilberto, Frattasi S., Pajukoski K. and Mogensen P. ,”
Single-user MIMO for LTE-A Uplink: Performance evaluation of OFDMA vs.
SC-FDMA”, Radio and Wireless Symposium, 2009. RWS '09. IEEE, pp. 304-307, 2009 .
[4] H. G. Myung, J. Lim and D. J. Goodman, “Single carrier FDMA for uplink wireless Transmission ,” Vehicular Technology Magazine, IEEE ,pp.30-38, Sep. 2006.
[5] H. G. Myung, J. Lim and D. J. Goodman, “Peak-to-Average Power Ratio of Single Carrier FDMA Signals with Pulse Shaping,” Proc. Of the IEEE PIMRC, pp.1-5 Sep.
2006.
[6] Mohammad A., Zekry A. and Newagy F.,”A Time Domain SLM for PAPR reduction in SC-FDMA Systems,” Global High Tech Congress on Electronics (GHTCE), 2006 IEEE, pp. 143-147, 2006.
[7] St´ephane Y. Le Goff, Samer S. Al-Samahi, Boon Kien Khoo, Charalampos C.
Tsimenidis, and Bayan S. Sharif, “Selected Mapping without Side Information for PAPR Reduction in OFDM,” IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, VOL. 8, NO. 7, JULY 2009
[8] Lu Xuan-min, WangYu-shiand ZhouYa-jian,“An Improved SLM Algorithm for PAPR Reduction in OFDM,” Information Engineering (ICIE), 2010 WASE International, pp.69-72, 2010.
[9] Wang Wen-Bo and Zheng Kan, “The OFDM Technique of broadband wireless,” Beijing, Post and Telecom press, 2003.
[10] Miao Wang, Zhangdui Zhong, Qiuyan Liu, ”Resource Allocation for SC-FDMA in LTE Uplink,” Service Operations, Logistics, and Informatics (SOLI), 2011 IEEE International Conference, pp.601-604, 2011.
49
[11] Shibata K. , Webber J., Nishimura T., Ohgane T., Ogawa, Y. ,” Blind Detection of Partial Transmit Sequence in a Coded OFDM System”, Global Telecommunications Conference, 2009. GLOBECOM 2009. IEEE, pp.1-6
[12] Robert J. Baxley and G. Tong Zhou,” Comparing Selected Mapping and Partial Transmit Sequence for PAR Reduction”, IEEE TRANSACTIONS ON BROADCASTING, VOL.
53, NO. 4, DECEMBER 2007
[13] Leonard J. Cimini, Jr. and Nelson R. Sollenberger, “Peak-to-Average Power Ratio Reduction of an OFDM Signal Using Partial Transmit Sequences” IEEE
COMMUNICATIONS LETTERS, VOL. 4, NO. 3, MARCH 2000
[14] Keunyoung K., Hoon, K., Youngnam, H. and Seong-Lyun, K., “Iterative and greedy resource allocation inan uplink OFDMA system” in proc. 15th IEEE International
Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC, Barcelona, Spain,September, 2004, pp. 2377-2381.
[15] Junsung Lim ,Myung, H.G. ; Kyungjin Oh ; Goodman, D.,” Channel-Dependent Scheduling of Uplink Single Carrier FDMA Systems”, Vehicular Technology Conference, 2006. VTC-2006 Fall. 2006 IEEE 64th
[16] Hyung G. Myung and David J. Goodman, Single Carrier FDMA A New Air Interface for Long Term Evolution, first edition, John Wiley & Sons, Ltd., United Kingdom, 2008.