|
Algorithm 1 Geo-location information aided spectrum sensing. |
Input:, , , WFPD. % for the q-th SUE, Output:.
-
1:
train the q-th T1-SVM classifier via
-
2:
train the q-th T2-SVM classifier via
-
3:
the q-th SUE estimates K TOAs (fingerprint) from the K BSs over
-
4:
if and it satisfies (15) then
-
5:
the estimated fingerprint best matches the -th WFP in the WFPD
-
6:
the q-th SUE is located in the -th grid
-
7:
else return to step 2
-
8:
end if
-
9:
the q-th SUE reads the labels in the -th WFP and obtain
-
10:
-
11:
ifthen
-
12:
if
then
-
13:
if
then
-
14:
output
-
15:
Exit to step 33
-
16:
end if
-
17:
end if
-
18:
end if
-
19:
if or
then the q-th SUE
-
20:
% single-SUE prediction
-
21:
obtains the sensing observation over the LFB
-
22:
predicts the state of the PUTs as via and the trained T1-SVM classifier
-
23:
predicts as via and the trained T2-SVM classifier
-
24:
if Joint-SUE prediction is required then
-
25:
reads received signals of the grids that neighbor the -th grid
-
26:
predicts independently, via single-SUE prediction
-
27:
obtains by HDF voting based on
-
28:
end if
-
29:
update in the -th WFP
-
30:
clear and restart in the -th WFP
-
31:
end if
-
32:
output
-
33:
if A new sensing request occurs then return to step 2
-
35:
Exit
|