Final comment by author

After a review of the previous exchanges of opinions routed through the Editor's good offices, the author of the original paper S.S. Kamal [ibid., vol. COM-27, pp. 857-868, June 1979] hopes that the following brief points may serve to clarify his position, and resolve the differences in interpretation. 1) In the original paper the definition of a homogeneous population of identical terminals was meant to convey uniformity in arrival statistics and steady-state network behavior. Such tyrminology is common throughout the literature, specifically, in references [4]-[8]. 2) Realizing this, the probability of a terminal transmitting anything in a given time slot is the sum of the probability that it transmits a new packet (i.e., one that has not been transmitted before, but not one that necessarily arrived to the buffer in the previous slot) plus the probability it transmits a previously collided packet. Thus, this probability of transmission in a slot is (/spl sigma/ + p). 3) A terminal can be unblocked, and with a buffer that has several packets. This occurs immediately after a blocked terminal frees itself (during which several arriving packets are forced to queue until serviced).