TensorRT how to implement the LRN layer with within_channel using iplugin?

In Tensorrt document, the LRN layer only supports cross channel. Why it does not support within_channel? If I want to implement the LRN layer with within_channel using Iplugin API, can you give me some suggestions?
Thanks!

Hi,

For better performance, different normalization axis require different implmentation.
In TensorRT 3, only cross-channel LRN is available.

If you want to implement it with Plugin API, here are some tutorial for your reference:
Native sample:
/usr/src/tensorrt/samples/samplePlugin/samplePlugin.cpp
/usr/src/tensorrt/samples/sampleCharRNN/sampleCharRNN.cpp
/usr/src/tensorrt/samples/sampleFasterRCNN/sampleFasterRCNN.cpp

TX2 sample:
https://github.com/AastaNV/Face-Recognition

Thanks.