Latency is caused by several phenomena, for instance errors at various layers, queueing, channel occupation and more. Techniques used to tackle these phenomena deeply affects latency, for example using UDP reduces latency, same for using 802.11n instead of 802.11g. Simulink provides the relevant blocksets to simulate a vast series of scenarios, but these blocks are usually black boxes you cannot modify. For example, if you want to try a new IP queueing method you may use the TCP/IP blocksets, but if you want to work on a layer 2 channel sense policy you must work with blocksets simulating the physical layers directly.

Dear Vinay,

the best way to evaluate the latency is to model the transmitter , the route of the signal which may be modeled as a core , and two links between the two terminals. After the modelling you can input some packets of the data and see their latency time and transmission rate.

The times measured are those of the implementation times by the used computing machine. They are only scaled values of the real times and they can be used as a relative time measures.

Best wishes