The Block Ack session is negotiated per Traffic Identifier (TID). Hence, in essence if a Block Ack session is negotiated for each TID – then
- The Block Ack bitmap for each of the sessions need to be saved.
- Also, the packets that are awaiting an Ack via the Block Ack frame need to be saved in memory until an Ack is received for the frame.
The above two requirements would consume a huge amount of memory in a WLAN station, particularly in an Access Point which would need to save Block Ack Bitmap and packets for Multiple stations.
To address the memory issue, the 802.11 standards organization suggested two methods of bitmap storage. They are
- Partial State Information
- Full State Information
Partial State Information
In Partial State Information, the Block Ack bitmap for a particular TID is saved only for the current Transmit Opportunity (TXOP). That is, when the TXOP finishes for the TID, then the Bitmap memory is also cleared out. Only certain information required for the movement of the Bitmap sliding window need to be retained. The packets which have been acked can be also removed from memory. The bitmap memory can be re-used for another TID in another TXOP.
The Partial State Information implies
- The Block Ack frame should be received in the current TXOP session.
- Since the bitmap memory is cleared out after the TXOP, it can only be used in an Immediate Block Ack method
Full State Information
In the case of Full State Information for a Block Ack Bitmap, The WLAN Station would need to save the bitmap context of the different block ack sessions fully. The Full State Information is extremely memory intensive and hence very difficult to implement.
- The implementation of a Full State Information allows “Delayed Block Ack”
- Certain Implementations might allow for full state information being saved for a particular TID.
Full state Information due to its memory constraints is very rarely implemented. It is also for the same reason delayed block ack is also rarely implemented.