Project #4: Design and implement an embedded, stand-alone QNX Neutrino program to simulate the workflow in a typical banking environment (single queue, multiple server queuing problem.)

Analysis: Customers: Customers enter the bank to transact business on a regular basis. Each new customer arrives every one to four minutes, based on a uniform random distribution. Each new customer enters a single queue of all customers. Tellers: Three tellers are available to service customers in the queue. As tellers become available, customers leave the queue, approach the teller and transact their business. Each customer requires between 30 seconds and 6 minutes to complete their transaction with the teller. The time required for each transaction is based on a uniform random distribution. Bank: The bank is open for business between the hours of 9:00am and 4:00pm. Customers begin entering when the bank opens in the morning, and stop entering when the bank closes in the afternoon. Customers in the queue at closing time remain in the queue until tellers are available to complete their transactions. Metrics: To monitor the performance of the business, metrics are gathered and reported at the end of the day. The metrics are:  the total number of customers serviced during the day,  the average time each customer spends waiting in the queue,  the average time each customer spends with the teller,  the average time tellers wait for customers,  the maximum customer wait time in the queue,  the maximum wait time for tellers waiting for customers,  the maximum transaction time for the tellers, and  the maximum depth of the queue.

Design Constraints: Each customer is represented as an application task (i.e. TCB). Each teller is modeled as an independent thread. The simulation parameters are variable and assigned at program startup. The simulation time is scaled such that 100 milliseconds of absolute clock time represents 1 minute of simulation clock time. The output should be presented in simulation clock time.