Stream Arbitration: Towards Efficient Bandwidth Utilization for Emerging On-Chip Interconnects