Thursday, 13 January 2011

RAN mechanisms to avoid CN overload due to MTC

Machine-to-Machine (M2M) is the future and Machine-type communications (MTC) will be very important once we have billions of connected devices. I have talked in the past about the 50 Billion connected devices by 2050 and the Internet of Things.

One of the challenges of today's networks is to handle this additional signalling traffic due to MTC. One of the very important topics being discussed in 3GPP RAN meetings is 'RAN mechanisms to avoid CN overload due to MTC'. Even though it has not been finalised, its interesting to see the direction in which things are moving.

The above figure from R2-106188 shows that an extended wait time could be added in the RRC Connection Reject/Release message in case if the eNodeB is overloaded. The device can reattempt the connection once the wait time has expired.

In R2-110462, another approach is shown where Core Network (CN) is overloaded. Here a NAS Request message is sent with delay tolerant indicator a.k.a. low priority indicator. If the CN is overloaded then it can reject the request with a backoff timer. Another approach would be to send this info to the eNodeB that can do a RRC Connection Reject when new connection request is received.

All Documents from 3GPP RAN2 #72-bis are available here. Search for NIMTC for M2M related and overload related docs.

No comments: