• 沒有找到結果。

Numerical Results and Discussion

Numerical performance of the proposed algorithms are presented in this section. We

0 100 200 300 400 500 600

Figure 3.2: The probability density function of the user locations; r0 = 150 m.

consider a network with four (M = 4) MS user nodes that are random distributed within a 120-degree section of the 600-meter radius circle centered at the BS. The probability density function (pdf) of the location is given by [11]

P = r40

3.3. Each transmitted signal experiences attenuation with a path loss exponent value of 3.5 and, in any direct or relay link, each subcarrier suffers from independent Rayleigh fading. For the convenience of comparison, we normalized the link gain with respect to the worst-case gain corresponding to the longest link distance. We assume that Ri = R = 128 ∀ i and each MS user is given 32 subcarriers so that N = 32 × 4 = 128.

105 simulation runs were performed to estimate the performance.

−3000 −200 −100 0 100 200 300 400 500 600 100

200 300 400 500 600

Figure 3.3: The probability density function of the user locations; r0 = 150 m.

Define the energy reduction ratio as the ratio between the total energy required to transmit a fixed amount of bits with a cooperative relay and that without a relay. In Fig.

3.4 we examine the influence of the priority threshold γ on the energy reduction ratio performance of our algorithms when the designed BER is (10−3). As expected, Algorithm B consistently outperforms Algorithm A for all thresholds. The reason is obvious: in Algorithm B, a source node is allowed to have multiple cooperative relay nodes, each is responsible for relaying data carried by certain subcarriers, and one can select the best link for every subcarrier. On the other hand, for Algorithm A, a source node can have at most one relay node which might have some good link quality in some subcarriers but not all of them. Hence Algorithm B enjoys a substantial performance gain at the cost of marginal complexity increase. For both algorithms, the optimal threshold γopt is about 1.4. Hence we use this value for subsequent simulations. The average energy reduction ratio performance of the two proposed algorithms is shown in Fig. 3.5. Similar to the previous figure, Algorithm B yields much more energy reduction than Algorithm A does

1 1.1 1.2 1.3 1.4 1.5 43.5

46 48.5 51 53.5 56 58.5 61 63.5 66 68.5

priority threshold value

percentage of energy consumption

Algorithm A Algorithm B

Figure 3.4: Energy reduction ratio performance as a function of the priority threshold γ with target BER = 10−3.

for all BER specification within the range [10−5, 10−1]. Furthermore, we find that the percentage of energy reduction is almost independent on the BER for both algorithms.

10−5 10−4

10−3 10−2

10−1 46 48 50 52 54 56 58 60 62 64 66 68

BER

percentage of energy consumption

Algorithm A Algorithm B

Figure 3.5: Energy reduction ratio performance of the proposed cooperative transmission schemes.

Table 3.1: Power Minimization Algorithm A (PMA).

if the user of hk is in assignment then k is assigned to it.

else

k is assigned to the other users in order of channel gains until every user gets enough subcarriers.

end end end

Step 5: Use equations (3.3) and (3.7) to complete bit-loading process and calculate the corresponding energy consumption.

Table 3.2: Power Minimization Algorithm B (PMB).

Step 1: for i = 1:M for k = 1:N

value(i; k) = max³

maxjhe(i, j; k),|hs(i;k)|2 ´ node(i; k) = arg

³

maxjhe(i, j; k),|hs(i;k)|2

´ end

end

Step 2: for i = 1:M if ηi ≥ 1

best − node = node end

end

The Step 3, Step 4 and Step 5 are the same as Algorithm A in Table I.

Chapter 4

Power, Minimum Rate and QoS

Constrained Sum Rate and Fairness Index Maximizing Resource

Allocation Schemes

In this chapter, we present resource (power, subcarriers and relays) allocation schemes that maximizes a fairness index and the sum rate with minimum rate and total power constraints for multiple-relay networks. By including cooperative nodes as part of the radio resources and taking into account the fairness issue, we propose two suboptimal algorithms that assign power, subcarriers and cooperative relay stations to a group of MS’s to meet their QoS and minimum rate requirements. It was shown that the sum rate of a multiuser OFDM system is maximized when each subcarrier is as-signed to the one which has the best channel condition. The total transmit power is then distributed over the subcarriers via a water-filling algorithm.

4.1 System Description and Basic Assumptions

We consider an N-subcarrier OFDMA-based cooperative communication network as that depicted in Fig. 4.1 in which there are M fixed relay nodes, K MS’ randomly dis-tributed within a cell centered at a BS. Assume that uplink channel state information

Figure 4.1: A cooperative communication network with multiple source and relay nodes and a single destination node.

is perfectly known to the BS which also knows the minimum rate and QoS (bit error rate) requirements of each source MS. The BS, acts as a central control device, will carry out all resource allocation operations, including collecting link information, ap-propriating resources, and informing MS’ about their assigned resources. Similar to the conventional relay-based cooperative communication systems, we assume a two-phase (time-slot) transmission scheme with perfect timing synchronization among all network users. Each subcarrier suffers from slow Rayleigh fading so that there is no change of the channel state during a two-phase period. A data stream from a source user must be carried by the same subcarrier no matter it is transmitted by a source node or a relay node.

The transmission pattern is half-duplex such that an MS transmits while the relay and the BS listen (receive) in the first time slot. In the second phase, the relay stations transmit to the BS while the source MS’ send new data packets via direct links without

relaying. This transmission protocol was discussed in [8] and was shown to be more throughput-efficient than the conventional protocol with which a source MS remains idle in the second phase. Both the decode-and-forward (DF) and amplify-and forward cooperative relay scheme are considered and the maximum-ratio-combining detector is employed by the destination (BS) node, assuming perfect decoding at the relays.

相關文件