The present investigation studies a discrete time single
server queue with both positive and negative arrival streams in
accordance with removal of the customer from the end (RCE)-in
immune and immune service killing policy. This study is a
generalization of the queue with negative customers, wherein only
positive customers need a service and negative customers arriving to
the system can kill the already present positive customers from any
where in the queue, otherwise get lost. The concept of both in-immune
and immune service killing are taken into consideration. According to
the in-immune killing policy, the negative customer is allowed to kill
the most recent positive customer inspite of whether it is in service or
not, while the immune service killing discipline suggests that the
customer currently being served is immune from killing by the
negative arrival. We analyze a queue with geometric arrivals of both
positive and negative customers for a finite capacity system. The
stationary probability distribution and other performance measures are
derived in terms of the generating functions. The results so obtained
are validated by the numerical method based on successive over
relaxation method (SOR). We have also employed the neurro fuzzy
approach for exhibiting the approximate results for various
performance measures.