1-6hit |
In this letter, we propose a simple but effective buffer management scheme to achieve fair bandwidth sharing with a FIFO scheduling algorithm, that is, Dynamic Detection and Expulsion (DDE). The DDE scheme dynamically detects buffer occupancy and then precisely expels resided packets on demand through simple comparisons. Simulation results under various traffic conditions show that DDE can arrive at more robust and better fairness, and lower implementation complexity than that of a well-known Pushout (PO) scheme.
Fawnizu Azmadi HUSSIN Tomokazu YONEDA Hideo FUJIWARA
Current NoC test scheduling methodologies in the literature are based on a dedicated path approach; a physical path through the NoC routers and interconnects are allocated for the transportation of test data from an external tester to a single core during the whole duration of the core test. This approach unnecessarily limits test concurrency of the embedded cores because a physical channel bandwidth is typically larger than the scan rate of any core-under-test. We are proposing a bandwidth sharing approach that divides the physical channel bandwidth into multiple smaller virtual channel bandwidths. The test scheduling is performed under the objective of co-optimizing the wrapper area cost and the resulting test application time using two complementary NoC wrappers. Experimental results showed that the area overhead can be optimized (to an extent) without compromising the test application time. Compared to other NoC scheduling approaches based on dedicated paths, our bandwidth sharing approach can reduce the test application time by up to 75.4%.
Seung Hyong RHEE Takis KONSTANTOPOULOS
A virtual private network (VPN) service is likely to be used by customers as a replacement for networks constructed using private lines, and thus its functionality should include the performance guarantee provided to those customers. To provide guaranteed services, the network provider allocates appropriate capacities to multiple virtual backbone networks such that the underlying network can be shared among them. As VPN users are demanding reliable and dynamic allocation of capacities, recently the capacity resizing approach has been considered as a cost efficient way of providing virtual network services. We propose a new scheme for dynamic allocation of virtual link capacities. The allocated capacities are adjusted dynamically according to the users' requests such that their capacities are increased in a fair manner and the total reservation does not overwhelm the underlying network. Depending on the network's status and allocation policy, a virtual link may increase or decrease its capacity, for example, for a monetary incentive. VPN users send control packets whenever they want to resize their capacities, and the network handles them in an efficient and fair way. The simulation and analytic results in this paper show that our scheme is simple and robust such that the users and the network communicate using simple control packets and the link capacities are allocated efficiently.
Yasuhiro SUZUKI Hideaki HARADA
A new optical access system based on the synchronous transfer mode - passive double star system has been developed to provide high-speed LAN-like access. It uses a shared-band method that enables multiple users to efficiently share a single bandwidth of up to 10 Mb/s and a grouping function that divides the access network into several logical networks, each of which can provide a virtual LAN to users. This paper describes an information model and a framework for configuration management and fault management and discusses the requirements for element management, which involves data-link establishment, logical group management, and testing. Element management mainly requires appropriate remote handling on data cards installed in each optical network unit on user premises. A method is proposed that satisfies these requirements. With this method, the element operations system can provide the required operational functionality.
Kohei NAKAI Eiji OKI Naoaki YAMANAKA
This paper proposes a 3-stage ATM switch architecture that uses optical WDM (wavelength division multiplexing) grouped links and dynamic bandwidth sharing. The proposed architecture has two features. The first is the use of WDM technology which makes the number of cables used in the system proportional to system size. The second is the use of dynamic bandwidth sharing among WDM grouped links. This prevents the statistical multiplexing gain offered by WDM from falling even if switching system becomes large. A performance evaluation confirms the scaleability and cost-effectiveness of the proposed architecture. It is scaleable in terms of the number of cables and admissible load. We show how the appropriate wavelength signal speed can be determined to implement the switch in a cost-effective manner. Therefore, the proposed architecture will suit future high-speed multimedia ATM networks.
Kohei NAKAI Eiji OKI Naoaki YAMANAKA
This paper proposes a 3-stage ATM switch architecture that uses optical WDM (wavelength division multiplexing) grouped links and dynamic bandwidth sharing. The proposed architecture has two features. The first is the use of WDM technology which makes the number of cables used in the system proportional to system size. The second is the use of dynamic bandwidth sharing among WDM grouped links. This prevents the statistical multiplexing gain offered by WDM from falling even if switching system becomes large. A performance evaluation confirms the scaleability and cost-effectiveness of the proposed architecture. It is scaleable in terms of the number of cables and admissible load. We show how the appropriate wavelength signal speed can be determined to implement the switch in a cost-effective manner. Therefore, the proposed architecture will suit future high-speed multimedia ATM networks.