# Job Sequencing using Branch and Bound

Job Sequencing using Branch and Bound: Given n jobs with profit, execution time, and deadline, achieve the schedule which maximizes the profit.

We will solve the problem using the FIFO branch and bound with the variable tuple and fixed tuple representations. Each job i is represented by the tuple (Pi, di, ti), where Pi, di, and ti represent profit, deadline, and execution time associated with job i. If job i is completed on or before its deadline, profit Pi is earned. But if the job i finish after its deadline, then a penalty Pi will incur. The brute force method finds 2n schedules for n jobs and finds the best from them. Branch and bound is a more efficient way of finding the optimal schedule.

## Examples on Job Sequencing using Branch and Bound

Example: Select optimal subset J with an optimal penalty for the following data. What will be the penalty corresponding to the optimal solution?

Solution:

The cost function for node x in state space tree for job sequencing problem is defined as,

$\hat{c}(x) = \sum_{i<m ; & ; i notin S_x} P_i$

Where, m = max{i  |  i ∈ Sx }

Sx is the subset of jobs selected at node x.

Upper bound u(x) for node x is defined as,

$u(x) = sum_{i notin S_x} P_i$

Computation of cost and bound for each node using variable tuple size is shown in the following table.

State-space tree for variable tuple formulation and fixed tuple formulation is depicted in Fig. (a) and Fig. (b), respectively Additional Reading: Read More