Channel State Prediction in Cognitive Radio, Part II: Single-User Prediction Zhe Chen, Nan Guo, Zhen Hu, and Robert C. Qiu

Presenter:

Zhe Chen

Department of Electrical and Computer Engineering Center for Manufacturing Research Tennessee Technological University Cookeville, Tennessee 38505 Email: [email protected]

Outline ■ Introduction ■ Channel State Prediction Using Modified Hidden Markov Model ■ Measurement of Wi-Fi Signals ■ Performance Evaluation ■ Conclusion

2

3/10/2011

Overview ■ The response delay has a negative impact on the accuracy of spectrum sensing, which is the cornerstone of cognitive radio. ■ In this paper, single-user channel state prediction is proposed to minimize the negative impact of the response delays. Specifically, a modified hidden Markov model (HMM)-based single-secondary-user (single-SU) prediction is proposed. ■ Real-world Wi-Fi signals are recorded and employed to evaluate the performance of the proposed approach.

3

3/10/2011

Outline ■ Introduction ■ Channel State Prediction Using Modified Hidden Markov Model ■ Measurement of Wi-Fi Signals ■ Performance Evaluation ■ Conclusion

4

3/10/2011

Hidden Markov Model (HMM) ■ An HMM is defined by a tuple □ □ □ □ □ □ □ □ □

- initial state probability vector - state transition probability matrix - emission probability matrix - the number of states - the states - the state at time , - the number of possible observation values - observation space - the observation value at time ,

5

3/10/2011

A Tailored Viterbi Algorithm for HMM ■ Given a parameter tuple and a sequence of observation values , the state sequence that is most likely to have generated the input sequence and the likelihood probability can be calculated using a tailored Viterbi algorithm:

- calculated likelihood probability - the estimated state at time 6

3/10/2011

Proposed Single-SU Prediction Approach ■ Proposed architecture for channel state prediction based on a modified HMM

■ Illustration of channel state prediction

7

3/10/2011

Proposed Modified HMM

■ ■

is the span of prediction, in the unit of time slots is the verification delay, in the unit of time slots

8

3/10/2011

Training ■ Training – to obtain for the proposed modified HMM ■ HMM training algorithm, like Baum-Welch algorithm, is not required in the proposed approach. ■ Instead, the parameters of the modified HMM are obtained through a simple statistical process over training sequences proposed as below:

9

- length of training sequence

3/10/2011

Prediction ■ Then a one-step prediction for the channel state X-slot ahead can be performed. ■ There are two methods for the proposed one-step prediction: □

□

- uses

- uses

and

and

for the prediction

for the prediction

10

3/10/2011

Prediction ■ Using either method, predicted channel state and the corresponding likelihood probability can be calculated.

11

3/10/2011

Outline ■ Introduction ■ Channel State Prediction Using Modified Hidden Markov Model ■ Measurement of Wi-Fi Signals ■ Performance Evaluation ■ Conclusion

12

3/10/2011

Wi-Fi Signal Measurement 2.3 MB/s date rate

Four locations

Simultaneously

6.25 GS/s sampling rate

40-ms duration

13

3/10/2011

Measured Wi-Fi Signals in Time-Domain

The weakest because of NLOS

The strongest and clearest due to the shortest propagation 14

3/10/2011

Outline ■ Introduction ■ Channel State Prediction Using Modified Hidden Markov Model ■ Measurement of Wi-Fi Signals ■ Performance Evaluation ■ Conclusion

15

3/10/2011

Performance Evaluation ■ The measured Wi-Fi signals from channel 1, 2, and 3 are fed to three independent SUs for prediction. ■ The measured Wi-Fi signal from channel 4 is served as an indicator of channel states and fed to all the three SUs for reference. ■ Set the length of time slot to 20 μs, and the duration of the spectrum sensing phase of a time slot to 4 μs. ■ Observation values: frequency-domain data from 2.418 GHz frequency tone after scalar quantization. ■ M, N, and Q for the single-SU prediction are set to 288, 2, and 1, respectively.

16

3/10/2011

Performance Evaluation ■ The proposed method is employed. ■ Reference approach: 1-nearest neighbor (1-NN) (X-slot ahead)

■ Two metrics □ □

- probability of detection - probability of false alarm

17

3/10/2011

Performance of Single-SU Prediction Proposed approach is better!

(channel #1)

(channel #2)

18

(channel #3)

3/10/2011

Outline ■ Introduction ■ Channel State Prediction Using Modified Hidden Markov Model ■ Measurement of Wi-Fi Signals ■ Performance Evaluation ■ Conclusion

19

3/10/2011

Conclusion ■ The response delays have been taken into account in designing a strategy for channel state prediction, and the strategy has been tested using real-world Wi-Fi signals recorded at four locations simultaneously. ■ An approach of single-SU prediction based on modified HMM has been proposed and tested. ■ Evaluation results confirm that the proposed single-SU prediction approach outperforms the 1-NN prediction approach, where the former asks for insignificant complexity increase and it does not need any thresholds. 20

3/10/2011

Acknowledgement ■ Grants from □ National Science Foundation □ Office of Naval Research

21

3/10/2011

Thank you!

22

3/10/2011