Cooperative Communications in Random Access Networks - Adaptive Signal Processing in Wireless Communications - page 327

and the corresponding rate indicated in the helper's table. When the helper receives the
message, it will check whether the rate indicated by the source is achievable. Inconsisten-
cies between the helper table and the actual achievable rate occur when the environment
varies and the entries in the table become out of date. If the helper is able to cooperate, it
will respond with an HTS message to notify the source and the destination that coopera-
tion is available and reserve the channel from its neighboring users. On the other hand,
if the relay node is busy (due to its own transmission or because of the transmission of
a neighbor), or it cannot achieve the rate requested by the source node, it will simply
remain silent, and after a certain time-out period expires, the destination node will pro-
ceed and reply with a CTS message to reserve the channel for direct transmission.
As shown above, the concept of cooperation can be easily incorporated into the IEEE
802.11 MAC with little changes to the original design. Several enhancement meth-
ods are also given in [15] to reduce the transmission overhead and increase the system
throughput.

11.5 Cooperation-Enhanced Collision

Resolution Methods

The essence of random access protocols is to enable users to access the channel in a
distributed fashion and deal with the collision afterwards. In the cooperative MAC
protocols introduced above, the advantages of relaying are exploited to combat fading in
the wireless channel, allowing the users that experience bad channels to be helped by the
users that experience good channels. However, these methods do not exploit the coop-
erative advantages in resolving the interference or collision at the destination. In this
and the following sections, we show that, with cooperation among users and appropriate
signal processing techniques at the destination, the mixture of signals that is received
during collision can be utilized to enhance the detection performance. In particular, we
describe in this section the system, proposed first by Tsatsanis et al. in [18] and then by
Lin and Petropulu in [17], that utilizes the variations in channel gains of the multiple
relaying paths to perform signal separation when collision occurs.
Consider a network of N nodes transmitting to an access point, similar to that shown
in Figure 11.2 . Conventionally, when more than one user is transmitting in the same
time slot, a mixture of signals will be received at the destination, causing strong inter-
ference among each other. In this case, it is likely that no packet will be received suc-
cessfully, resulting in the so-called collision . The mixture of signals is usually discarded
at the destination, and no information is extracted from the signals, which is clearly a
waste of energy and bandwidth.
Due to this reason, Tsatsanis et al. [18] proposed a network-assisted collision resolu-
tion method. In this method, the destination records the mixture of signals whenever
a collision occurs and enables the same set of colliding users to continue transmitting
simultaneously in a certain number of subsequent time slots. Specifically, let x i [ n ] =
[ x i ,1 [ n ], … , x i,M [ n ]] T be the M -bit message transmitted by user i in the n th time slot, and let
I[ n ] ⊂ {1, 2, …, N } be the set of users that are transmitting in the n th time slot. A collision