In this work, a heterogeneous cellular network with caching enabled at the small cell aggregators is analyzed. The caching technology is enabled to help offload some of the popular traffic requests onto the cache. In this framework, a network congestion game is formulated to study the evolution of traffic when every user as an independent decision maker decides its choice of communication, i.e, via the macro cell or the caching enabled small cells. To study the effect of the cache on the network delay, a numerical simulation is implemented. The file popularity is modeled using the well known Zipf-distribution and the network delay is studied as a function of the cache buffer size and network traffic. In summary, these results can be utilized by network operators to deploy infrastructure efficiently.