18 research outputs found
Enhanced response algorithm for spurious TCP timeout(ER-SRTO)
No full textEifel and F-RTO have been reported representative solutions for spurious TCP timeout. However, these algorithms do not work well in case that delay spike and packet loss occur simultaneously. In this paper, we propose ER-SRTO algorithm which detects the packet loss during delay spike and conducts an efficient recovery in the RTO procedure based on ACK sequence number of retransmission packet. Proposed scheme recovers the lost packet without invoking additional loss recovery procedure. Simulation result shows that proposed algorithm outperforms previous ones
Timer and sequence based packet loss detection scheme for efficient selective retransmission in DCCP
No full textData Congestion Control Protocol (DCCP) is new transport protocol for multimedia applications. Although the standard DCCP does not provide retransmission scheme, many research works have proposed selective retransmission methods for DCCP to improve QoS of applications. However, these researches do not consider packet loss detection scheme although existing packet loss detection schemes are not appropriate for efficient retransmission. This paper proposes a new packet loss detection method for efficient selective retransmission in DCCP. The proposed scheme detects packet loss by using timer and out of sequence packets. Simulation results show that the proposed method improves the performance of retransmission by detecting packet loss more quickly and accurately
Análisis de priorización de requerimientos en un servicio de atención a clientes
No full textMemoria para optar al título de Ingeniero Civil IndustrialLa priorización de requerimientos es una práctica recurrente en diversas firmas de servicio por lo que tiene gran interés estudiar su conveniencia y aplicabilidad. Algunas investigaciones sobre esta materia sugieren que los administradores a cargo de planificar la atención a clientes deberían implementar esta política, pues en general, tiene un efecto positivo sobre la rentabilidad de los negocios y en particular sobre el retorno obtenido por las ventas. Pese a esto, es difícil establecer si su implementación ayudará a generar los beneficios suficientes que permitan justificar su ejecución, además, se hace necesario abordar cómo incorporar esta práctica en la planificación de operaciones de un servicio específico. Bajo este contexto, se ha desarrollado una metodología que permite incluir información de los clientes de una firma para priorizar la ejecución de sus requerimientos de atención. De esta forma se han caracterizado algunos escenarios de demanda que permiten dilucidar cuándo se perciben los mayores beneficios asociados a esta práctica y qué tipo de información de los clientes puede ser incorporada como base. La metodología mencionada no solo ayuda a estimar de forma general el ahorro de recursos adquiridos al priorizar la atención, también permite caracterizar ciertos indicadores globales de servicio obtenidos al simular la atención de una cartera de clientes. Mediante este trabajo se ha determinado que, en el mejor escenario caracterizado, la firma de servicios podría ahorrar CLP3.690 por cliente flexibilizado. Se ha mostrado que la política estudiada es económicamente factible de implementar en ciertos casos, pues se produce una liberación de recursos que pueden ser utilizados en la diferenciación del servicio ofrecido a los clientes. Pese a lo anterior, su viabilidad práctica depende de que las condiciones específicas del negocio permitan proveer el servicio a la cartera restante de clientes a un menor costo del que se provee en la actualidad
Enhanced TCP Congestion Control for Vertical Handover with the RTT inflation and the measured-RTT of the new network
No full textInterference Avoidance and Resource Allocation for OFDMA Downlink of Femtocells with Distributed Power Control
No full textOFDMA femtocells in the macrocellular network of which frequency reuse factor is I cause serious interference to macrocell users. while the femtocells improve the performance of indoor users In this letter, a novel downlink resource allocation algorithm for OFDMA femtocell networks is proposed to reduce interference between the macrocells and the femtocells This algorithm allocates femtocell subchannels to avoid interference to macrocell users in the femtocell coverage and minimizes the total transmission power of the femtocell to reduce the negative effect on the performance of the macrocell Simulation results ate provided to present the performance of the proposed algorith
Renewable Butanol Production via Catalytic Routes
No full textFluctuating crude oil price and global environmental problems such as global warming and climate change lead to growing demand for the production of renewable chemicals as petrochemical substitutes. Butanol is a nonpolar alcohol that is used in a large variety of consumer products and as an important industrial intermediate. Thus, the production of butanol from renewable resources (e.g., biomass and organic waste) has gained a great deal of attention from researchers. Although typical renewable butanol is produced via a fermentative route (i.e., acetone-butanol-ethanol (ABE) fermentation of biomass-derived sugars), the fermentative butanol production has disadvantages such as a low yield of butanol and the formation of byproducts, such as acetone and ethanol. To avoid the drawbacks, the production of renewable butanol via non-fermentative catalytic routes has been recently proposed. This review is aimed at providing an overview on three different emerging and promising catalytic routes from biomass/organic waste-derived chemicals to butanol. The first route involves the conversion of ethanol into butanol over metal and oxide catalysts. Volatile fatty acid can be a raw chemical for the production of butanol using porous materials and metal catalysts. In addition, biomass-derived syngas can be transformed to butanol on non-noble metal catalysts promoted by alkali metals. The prospect of catalytic renewable butanol production is also discussed
TCP-Friendly Retransmission Persistence Management for SR-ARQ Protocols
No full textThis letter proposes a new retransmission persistence management scheme for selective repeat automatic repeat request (SR-ARQ) By considering the overall traffic load that has to be managed by SR-ARQ, the proposed scheme arbitrates the retransmission persistence to prevent an abrupt delay increment due to excessive link-level local retransmissions. OPNET simulations show that SR-ARQ performs better with the proposed scheme than with a fixed value of retransmission persistence in terms of the throughput of transmission control protocol (TCP)
