IEEE 802 Study Group on BWA: MAC Task Group Input


(1) From: Allan Evans, 19 Jan. 1999

I recommend that we start listing requirements for the MAC, to determine the applicability of standardized or proprietary MACs to broadband wireless access.

As a start:

1. Support multiple services and multiple interworking functions to customer premises equipment. While it is beyond the scope of this effort to define interworking functions, it is necessary to list CPE interfaces and IWFs to understand what services the MAC layer must provide to various higher layers.

2. Support QoS including: Cell/Frame Loss Ratio Loss to provide equivalent performance to all fiber systems End to end delay and Cell Delay Variation / Packet Arrival Rate Variation to support voice and video services Maximum Burst Rate and Sustained Rate to enable service tiering and service level agreements

3. Support symmetric services to support delivery of telecom services to small/medium businesses. Asymmetric services should not be precluded as they will be required for residential users.

4. Scaleability to support high bit rates per channel, DS3 and higher.

After we agree on the requirements, then we should issue a call for proposals for MAC protocols. Submissions will likely include extensions to existing or developing standards as well as new proprietary schemes being opened for standardization. Leave it to the members of the task group to propose extensions to existing standards in response to the call for proposals.


(2) From: Marianna Goldhammer, 20 Jan. 1999

I would add:

5. Support for high efficiency in FDD ( frequency division duplex) operation

6. Support for high frequency reuse in TDD (time division duplex) operation

7. Support for packet switched opperation with time bounded and/or random access


(3) From: Allan Evans, 20 Jan. 1999

Specifying TDD or FDD is the charter of the PHY Group or the Coexistence Group. We should get a ruling on this by Roger. It would be premature for the MAC group to consider this.

I would like to clarify 7) to say the following:

7) Provide statistical multiplexing gain for packet switched services.


(4) From: Marianna Goldhammer, 27 Jan. 1999

According to my previous experience TDD vs. FDD is a dispute that will never end, so I think that we have to keep in mind the both possibilities.

Regarding the point 7, in order to make thinks clear, lets split it in 2 points:

7. Provide statistical multiplexing gain for packet switched services.

8. Provide time bounded services for time critical packet data (Voice over IP, Video over IP).


Roger Marks (r.b.marks@ieee.org)