I have no knowledge about the TKN implementation. However, I don´t agree with the interpretation of item 2. Because, when the backoff counter reach 0, the station can transmit immediately, i.e., in the next slot time. The station doesn´t wait for another AIFS interval.
> > According to section 9.9.1.3 in the IEEE 802.11e standard, > > one and only one of the following actions is allowed at slot boundaries: > - decrement of backoff timer, or > - initiate transmission, or > - backoff procedure due to internal collision, or > - do nothing. > > > So, > > 1. > Backoff counter is resumed one slot time before the AIFS expiration. > > 2. > However, since a single MAC operation per slot is permitted > (backoff decrement or packet transmission), > > when the backoff counter decrements to 0, the station cannot > transmit immediately, but has to wait for a further backoff > slot if the medium is idle, or a further AIFS expiration if > the medium is busy. > > > I think that 802.11e TKN code is a good implementation of IEEE 802.11e. > > However, it seems that the implementation conforms to the requirement > 1, but do not to the requirement 2. > > Am I correct? > > > ps. > The following paper can be a good guide. > > Giuseppe Bianchi, "Understanding 802.11e Contention-Based > Prioritization Mechanisms and Their Coexistence with Legacy 802.11 > Stations", IEEE Network, 2005 > >
