1.2.3 Multiple Access Channels
1.2.3 Multiple Access Channels
Multiaccess
Source 2 Destination 2
Channel
Source 3 Destination 3
The channels we have considered so far are called point-to-point channels where the received
signal at the destination node depends only on the signal transmitted by the source node and the
channel noise. A point-to-point channel is used exclusively by a single source-destination node
pair for communication. However, there are some channels called multiaccess or multiple access
channels which are shared by several source-destination node pairs for communication. In this
case, the received signal at a destination node depends on the signal transmitted by several source
nodes. Typically, it is the sum of the attenuated transmitted signals corrupted by channel noise.
Figure 1.5 illustrates a multiple access communication scenario between three source-destination
node pairs. The received signal at any destination depends on the signal transmitted by all three
source nodes. The simplest example of a multiple access channel is the wireless channel when
multiple source-destination node pairs located within each other’s transmission range use the
same frequency band to communicate. Multiple access communication does not necessarily
imply an equal number of sources and destinations. For example, in satellite communication a
single satellite may be communicating with multiple ground stations using the same frequency
band.
Multiple access channels are not always wireless. For example, consider a collection of nodes
which are located in close proximity to each other. Suppose each node in this collection needs
to communicate with every other node. One method to achieve this is to create a point-to-point
wired channel from each node to every other node. This is illustrated in Figure 1.6a for a
collection of eight nodes where the shaded circles represent the nodes and the edges represent
the wired channels. Such an approach does not scale well as the number of nodes increases. A
more cost-effective solution which is used in practice is to connect the nodes using a bus which is,
in this context, a wired multiple access channel as illustrated in Figure 1.6b. Any transmission
by a node is sent across the bus and is heard by all the nodes. The node which is the intended
receiver will demodulate the received signal and obtain the information destined for it.
There are two issues with the system shown in Figure 1.6b which are common to all multiple
access communication systems. The first one is addressing. The source node has to include the
identity or address of the intended receiver of the information along with the information itself.
This will enable all the nodes except the intended receiver to ignore the information and the
intended receiver to accept it. The address of the receiver or the destination address is trans-
mitted before the information bits as part of the header. The reason is that the receiving nodes
need not process the whole information bit string before ignoring the transmission. Typically
the receiving node needs to send back some information back to the source node. To enable this
the header also contains the address of the source node or the source address. Of course, we
have not discussed what exactly constitutes an address. The address structure is highly system