Algorithm 1 Federated stochastic gradient averaging learning

From: Federated stochastic gradient averaging ring homomorphism based learning for secure data aggregation in WSN

Input: sensing data ‘\(\:DP=\left\{{DP}_{1},\:{DP}_{2},\:\dots\:,{DP}_{N}\right\}\)’, routing paths ‘\(\:R=\left\{{R}_{1},\:{R}_{2},\:\dots\:,{R}_{M}\right\}\)’, sink node ‘\(\:S\)’, sensor node ‘\(\:{SN=\{SN}_{1},\:{SN}_{2},\:\dots\:,{SN}_{m}\}\)’

Output: Computationally-efficient aggregated learning results

1: Initialize ‘\(\:N\)’, monitoring area ‘\(\:Area\)’, ‘\(\:M\)’, ‘\(\:{E}_{0}\)’, communication radius ‘\(\:Rad\)’, ‘\(\:m\)’

2: Begin

3: For each sensing data ‘\(\:DP\)’ with routing paths ‘\(\:R\)’, monitoring area ‘\(\:Area\)’ and communication radius ‘\(\:Rad\)’

//sensed data packet learning for the corresponding sensor nodes in monitoring area within the communication radius

4: Formulate initial learning process as given in Eq. (1)

5: Update weight values as given in Eq. (2)

6: Obtain aggregated learning results as given in Eq. (3)

7: End for

8: End

Back to article page