Thesis (M.S. in Operations Research) Naval Postgraduate School, June 1997. / Thesis advisor, Arnold Buss. Includes bibliographical references (p. 115). Also available online.

International Telemetering Conference Proceedings / October 17-20, 1994 / Town & Country Hotel and Conference Center, San Diego, California / In recent years the extension of interactive simulation technology to involve simulators and live vehicles from geographically dispersed sites has produced a demand for high-bandwidth communication networks that can provide guaranteed quality of service (e.g., insured availability of bandwidth and upper bounds on end-to-end delay). This paper reviews the requirements distributed interactive simulation places on the communications infrastructure and describes the Defense Simulation Internet (DSI), a network developed by the Advanced Research Projects Agency to support distributed interactive simulations. Key features of the DSI are: more than 120 participating sites spanning Europe, the United States and Asia; use of a resource reservation mechanism to provide guaranteed quality of service; and support for communication between classified sites. Furthermore, the paper describes the internetworking protocols used in the DSI to provide guaranteed quality of service and to support transmission of classified communications. Other topics discussed in the paper are research efforts that anticipate increased load on the DSI and the relevance of the technology to the integration of the telemetry range and distributed interactive simulations.

International Telemetering Conference Proceedings / October 17-20, 1994 / Town & Country Hotel and Conference Center, San Diego, California / The Distributed Interactive Simulation (DIS) concept, since its inception, has been defined into three separate but distinct areas of service. • Viewing of data in the real-time environment. • Multiple range viewing and usage of"real-time data." • Problems with the sharing of information through DIS. This paper will discuss the DIS concept and some of the various methods available to display this data to users of the system.

International Telemetering Conference Proceedings / October 27-30, 1997 / Riviera Hotel and Convention Center, Las Vegas, Nevada / The U.S. Air Force has developed GPS-based instrumentation systems to support both test and training activities. In support of recent large-scale exercises, interfaces were developed to employ existing test and training assets in a synthetic battlefield. The writers propose exploration of similar approaches to overcome the challenge of developing a common approach to test and training instrumentation.

International Telemetering Conference Proceedings / October 28-31, 1996 / Town and Country Hotel and Convention Center, San Diego, California / This paper discusses Global Positioning System (GPS) Range Applications Joint Program Office (RAJPO) efforts to foster interoperability between airborne instrumentation, virtual simulators, and constructive simulations using Distributed Interactive Simulation (DIS). In the past, the testing and training communities developed separate airborne instrumentation systems primarily because available technology couldn't encompass both communities' requirements. As budgets get smaller, as requirements merge, and as technology advances, the separate systems can be used interoperably and possibly merged to meet common requirements. Using DIS to bridge the gap between the RAJPO test instrumentation system and the Air Combat Maneuvering Instrumentation (ACMI) training systems provides a defacto system-level interoperable interface while giving both communities the added benefits of interaction with the modeling and simulation world. The RAJPO leads the test community in using DIS. RAJPO instrumentation has already supported training exercises such as Roving Sands 95, Warfighter 95, and Combat Synthetic Test, Training, and Assessment Range (STTAR) and major tests such as the Joint Advanced Distributed Simulation (JADS) Joint Test and Evaluation (JT&E) program. Future efforts may include support of Warrior Flag 97 and upgrading the Nellis No-Drop Bomb Scoring Ranges. These exercises, combining the use of DIS and RAJPO instrumentation to date, demonstrate how a single airborne system can be used successfully to support both test and training requirements. The Air Combat Training System (ACTS) Program plans to build interoperability through DIS into existing and future ACMI systems. The RAJPO is committed to fostering interoperable airborne instrumentation systems as well as interfaces to virtual and constructive systems in the modeling and simulation world. This interoperability will provide a highly realistic combat training and test synthetic environment enhancing the military's ability to train its warfighters and test its advanced weapon systems.

International Telemetering Conference Proceedings / October 30-November 02, 1995 / Riviera Hotel, Las Vegas, Nevada / GPS user equipment has matured and is now available to support the use of live players in integrated ranges and simulations. P-code GPS provides true WGS-84 based coordinate information anywhere in the world at any time and to accuracies at the 5 ft (1s) level (demonstrated in high dynamic aircraft using differential P-code GPS). C/A code GPS shows lower accuracy and is especially vulnerable to multipath degradation over water. In supporting networked ranges with simulations, GPS is directly applicable to the dead reckoning requirements of the Distributed Interactive Simulation (DIS) community. DIS dead reckoning provides the capability of much reduced data rates in recovering TSPI information from platforms. The on-board state vector for an integrated GPS/Inertial Reference Unit provides accurate position, velocity and acceleration as well as attitude and attitude rate information so that dead reckoning thresholds can be both position and attitude driven. A simplified analysis is presented in the paper to derive dead reckoning update rates from the G loading levels of various player dynamics. Also, information is provided which results in word length requirements for GPS-based state vector information for transmission over minimum word length DIS Field Instrumentation Protocol Data Units (PDUs, which are the data block formats). The coordinate frame problem in use of GPS-based state vector information from fixed ranges is also addressed, showing that the use of a local geodetic frame is preferable to the use of an earth centered earth fixed frame, in that it is more efficient of network PDU word length.

Tradicionalmente, utiliza-se um servidor central para prover suporte a MMOGs (massively multiplayer online games, ou jogos online maciçamente multijogador), nos quais o número de participantes é da ordem de dezenas de milhares. Muitos trabalhos foram realizados com o intuito de criar um modelo de suporte completamente descentralizado, par-a-par, para este tipo de aplicação, minimizando o custo de manutenção da sua infraestrutura, mas algumas questões críticas persistem. Exemplos de problemas do modelo de suporte par-a-par são: vulnerabilidade a trapaça, sobrecarga da banda de envio dos pares e dificuldade para manter a consistência da simulação entre os diferentes participantes. Neste trabalho, é proposta a utilização de nodos de baixo custo geograficamente distribuídos, operando como um servidor distribuído de jogo. O modelo de distribuição proposto e alguns trabalhos relacionados também são apresentados. Para tratar o custo de comunicação imposto aos servidores, foi projetado aqui um novo refinamento para a técnica de gerenciamento de interesse, reduzindo significativamente a largura de banda necessária ao jogo. Foram realizadas simulações utilizando o simulador ns-2, comparando diferentes algoritmos de área de interesse. Os resultados demonstram que a nossa proposta é a que tem a menor utilização de largura de banda, com uma redução em 33,10% do tráfego máximo, e em 33,58% do tráfego médio, quando comparada com outros algoritmos de gerenciamento de interesse. Além disso, em uma arquitetura de servidor distribuído para MMOGs, com recursos heterogêneos, os nodos servidores podem ser facilmente sobrecarregados pela alta demanda dos jogadores por atualizações de estado. Neste trabalho, é proposto também um esquema de balanceamento de carga que utiliza o tráfego de rede como a carga a balancear entre os servidores e tem como objetivos principais: alocar a carga nos servidores proporcionalmente à capacidade de cada um e reduzir tanto quanto possível o overhead introduzido pela própria distribuição. O esquema de balanceamento é dividido em três fases: seleção local de servidores para participarem, o balanceamento em si e o posterior refinamento da divisão de carga. Quatro algoritmos foram propostos: ProGReGA, ProGReGA-KH, ProGReGA-KF e BFBCT. Destes, o ProGReGA foi o que introduziu o menor overhead de todos e o ProGReGA-KF foi o que se mostrou mais eficiente para reduzir o número de migrações de jogadores entre servidores. / Traditionally, a central server is used to provide support to MMOGs (massively multiplayer online games), where the number of participants is in the order of tens of thousands. Much work has been done trying to create a fully peer-to-peer model to support this kind of application, in order to minimize the maintenance cost of its infra-structure, but critical questions remain. Examples of the problems relative to peer-to-peer MMOG support systems are: vulnerability to cheating, overload of the upload links of the peers and difficulty to maintain consistency of the simulation among the participants. In this work, it is proposed the utilization of geographically distributed lower-cost nodes, working as a distributed server to the game. The distribution model and some related works are also presented. To address the communication cost imposed to the servers, we specify a novel refinement to the area of interest technique, significantly reducing the necessary bandwidth. Simulations have been made with ns-2, comparing different area of interest algorithms. The results show that our approach achieves the least bandwidth utilization, with a 33.10% maximum traffic reduction and 33.58% average traffic reduction, when compared to other area of interest algorithms. Besides, in a distributed MMOG server architecture, with heterogeneous resources, the server nodes may become easily overloaded by the high demand from the players for state updates. In this work, we also propose a load balancing scheme, which considers the network traffic as the load to balance between the servers, and it has the following main objectives: allocate load on the servers proportionally to the power of each one of them and reduce as much as possible the overhead introduced by the distribution itself. It is is divided in three phases: local selection of servers, balancing and refinement. Four algorithms were proposed: ProGReGA, ProGReGA-KH, ProGReGA-KF and BFBCT. From these, ProGReGA has proved to be the best for overhead reduction and ProGReGA-KF is the most suited for reducing player migrations between servers.

Tradicionalmente, utiliza-se um servidor central para prover suporte a MMOGs (massively multiplayer online games, ou jogos online maciçamente multijogador), nos quais o número de participantes é da ordem de dezenas de milhares. Muitos trabalhos foram realizados com o intuito de criar um modelo de suporte completamente descentralizado, par-a-par, para este tipo de aplicação, minimizando o custo de manutenção da sua infraestrutura, mas algumas questões críticas persistem. Exemplos de problemas do modelo de suporte par-a-par são: vulnerabilidade a trapaça, sobrecarga da banda de envio dos pares e dificuldade para manter a consistência da simulação entre os diferentes participantes. Neste trabalho, é proposta a utilização de nodos de baixo custo geograficamente distribuídos, operando como um servidor distribuído de jogo. O modelo de distribuição proposto e alguns trabalhos relacionados também são apresentados. Para tratar o custo de comunicação imposto aos servidores, foi projetado aqui um novo refinamento para a técnica de gerenciamento de interesse, reduzindo significativamente a largura de banda necessária ao jogo. Foram realizadas simulações utilizando o simulador ns-2, comparando diferentes algoritmos de área de interesse. Os resultados demonstram que a nossa proposta é a que tem a menor utilização de largura de banda, com uma redução em 33,10% do tráfego máximo, e em 33,58% do tráfego médio, quando comparada com outros algoritmos de gerenciamento de interesse. Além disso, em uma arquitetura de servidor distribuído para MMOGs, com recursos heterogêneos, os nodos servidores podem ser facilmente sobrecarregados pela alta demanda dos jogadores por atualizações de estado. Neste trabalho, é proposto também um esquema de balanceamento de carga que utiliza o tráfego de rede como a carga a balancear entre os servidores e tem como objetivos principais: alocar a carga nos servidores proporcionalmente à capacidade de cada um e reduzir tanto quanto possível o overhead introduzido pela própria distribuição. O esquema de balanceamento é dividido em três fases: seleção local de servidores para participarem, o balanceamento em si e o posterior refinamento da divisão de carga. Quatro algoritmos foram propostos: ProGReGA, ProGReGA-KH, ProGReGA-KF e BFBCT. Destes, o ProGReGA foi o que introduziu o menor overhead de todos e o ProGReGA-KF foi o que se mostrou mais eficiente para reduzir o número de migrações de jogadores entre servidores. / Traditionally, a central server is used to provide support to MMOGs (massively multiplayer online games), where the number of participants is in the order of tens of thousands. Much work has been done trying to create a fully peer-to-peer model to support this kind of application, in order to minimize the maintenance cost of its infra-structure, but critical questions remain. Examples of the problems relative to peer-to-peer MMOG support systems are: vulnerability to cheating, overload of the upload links of the peers and difficulty to maintain consistency of the simulation among the participants. In this work, it is proposed the utilization of geographically distributed lower-cost nodes, working as a distributed server to the game. The distribution model and some related works are also presented. To address the communication cost imposed to the servers, we specify a novel refinement to the area of interest technique, significantly reducing the necessary bandwidth. Simulations have been made with ns-2, comparing different area of interest algorithms. The results show that our approach achieves the least bandwidth utilization, with a 33.10% maximum traffic reduction and 33.58% average traffic reduction, when compared to other area of interest algorithms. Besides, in a distributed MMOG server architecture, with heterogeneous resources, the server nodes may become easily overloaded by the high demand from the players for state updates. In this work, we also propose a load balancing scheme, which considers the network traffic as the load to balance between the servers, and it has the following main objectives: allocate load on the servers proportionally to the power of each one of them and reduce as much as possible the overhead introduced by the distribution itself. It is is divided in three phases: local selection of servers, balancing and refinement. Four algorithms were proposed: ProGReGA, ProGReGA-KH, ProGReGA-KF and BFBCT. From these, ProGReGA has proved to be the best for overhead reduction and ProGReGA-KF is the most suited for reducing player migrations between servers.

Tradicionalmente, utiliza-se um servidor central para prover suporte a MMOGs (massively multiplayer online games, ou jogos online maciçamente multijogador), nos quais o número de participantes é da ordem de dezenas de milhares. Muitos trabalhos foram realizados com o intuito de criar um modelo de suporte completamente descentralizado, par-a-par, para este tipo de aplicação, minimizando o custo de manutenção da sua infraestrutura, mas algumas questões críticas persistem. Exemplos de problemas do modelo de suporte par-a-par são: vulnerabilidade a trapaça, sobrecarga da banda de envio dos pares e dificuldade para manter a consistência da simulação entre os diferentes participantes. Neste trabalho, é proposta a utilização de nodos de baixo custo geograficamente distribuídos, operando como um servidor distribuído de jogo. O modelo de distribuição proposto e alguns trabalhos relacionados também são apresentados. Para tratar o custo de comunicação imposto aos servidores, foi projetado aqui um novo refinamento para a técnica de gerenciamento de interesse, reduzindo significativamente a largura de banda necessária ao jogo. Foram realizadas simulações utilizando o simulador ns-2, comparando diferentes algoritmos de área de interesse. Os resultados demonstram que a nossa proposta é a que tem a menor utilização de largura de banda, com uma redução em 33,10% do tráfego máximo, e em 33,58% do tráfego médio, quando comparada com outros algoritmos de gerenciamento de interesse. Além disso, em uma arquitetura de servidor distribuído para MMOGs, com recursos heterogêneos, os nodos servidores podem ser facilmente sobrecarregados pela alta demanda dos jogadores por atualizações de estado. Neste trabalho, é proposto também um esquema de balanceamento de carga que utiliza o tráfego de rede como a carga a balancear entre os servidores e tem como objetivos principais: alocar a carga nos servidores proporcionalmente à capacidade de cada um e reduzir tanto quanto possível o overhead introduzido pela própria distribuição. O esquema de balanceamento é dividido em três fases: seleção local de servidores para participarem, o balanceamento em si e o posterior refinamento da divisão de carga. Quatro algoritmos foram propostos: ProGReGA, ProGReGA-KH, ProGReGA-KF e BFBCT. Destes, o ProGReGA foi o que introduziu o menor overhead de todos e o ProGReGA-KF foi o que se mostrou mais eficiente para reduzir o número de migrações de jogadores entre servidores. / Traditionally, a central server is used to provide support to MMOGs (massively multiplayer online games), where the number of participants is in the order of tens of thousands. Much work has been done trying to create a fully peer-to-peer model to support this kind of application, in order to minimize the maintenance cost of its infra-structure, but critical questions remain. Examples of the problems relative to peer-to-peer MMOG support systems are: vulnerability to cheating, overload of the upload links of the peers and difficulty to maintain consistency of the simulation among the participants. In this work, it is proposed the utilization of geographically distributed lower-cost nodes, working as a distributed server to the game. The distribution model and some related works are also presented. To address the communication cost imposed to the servers, we specify a novel refinement to the area of interest technique, significantly reducing the necessary bandwidth. Simulations have been made with ns-2, comparing different area of interest algorithms. The results show that our approach achieves the least bandwidth utilization, with a 33.10% maximum traffic reduction and 33.58% average traffic reduction, when compared to other area of interest algorithms. Besides, in a distributed MMOG server architecture, with heterogeneous resources, the server nodes may become easily overloaded by the high demand from the players for state updates. In this work, we also propose a load balancing scheme, which considers the network traffic as the load to balance between the servers, and it has the following main objectives: allocate load on the servers proportionally to the power of each one of them and reduce as much as possible the overhead introduced by the distribution itself. It is is divided in three phases: local selection of servers, balancing and refinement. Four algorithms were proposed: ProGReGA, ProGReGA-KH, ProGReGA-KF and BFBCT. From these, ProGReGA has proved to be the best for overhead reduction and ProGReGA-KF is the most suited for reducing player migrations between servers.

Maintaining coherence between the independent views of multiple participants at distributed locations is essential in an Embedded Simulation environment. Currently, the Distributed Interactive Simulation (DIS) protocol maintains coherence by broadcasting the entity state streams from each simulation station. In this dissertation, a novel alternative to DIS that replaces the transmitting sources with local sources is developed, validated, and assessed by analytical and experimental means. The proposed Concurrent Model approach reduces the communication burden to transmission of only synchronization and model-update messages. Necessary and sufficient conditions for the correctness of Concurrent Models in a discrete event simulation environment are established by developing Behavioral Congruence ¨B(EL, ER) and Temporal Congruence ¨T(t, ER) functions. They indicate model discrepancies with respect to the simulation time t, and the local and remote entity state streams EL and ER, respectively. Performance benefits were quantified in terms of the bandwidth reduction ratio BR=N/I obtained from the comparison of the OneSAF Testbed Semi-Automated Forces (OTBSAF) simulator under DIS requiring a total of N bits and a testbed modified for the Concurrent Model approach which required I bits. In the experiments conducted, a range of 100 d BR d 294 was obtained representing two orders of magnitude reduction in simulation traffic. Investigation showed that the models rely heavily on the priority data structure of the discrete event simulation and that performance of the overall simulation can be enhanced by an additional 6% by improving the queue management. A low run-time overhead, self-adapting storage policy called the Smart Priority Queue (SPQ) was developed and evaluated within the Concurrent Model. The proposed SPQ policies employ a lowcomplexity linear queue for near head activities and a rapid-indexing variable binwidth calendar queue for distant events. The SPQ configuration is determined by monitoring queue access behavior using cost scoring factors and then applying heuristics to adjust the organization of the underlying data structures. Results indicate that optimizing storage to the spatial distribution of queue access can decrease HOLD operation cost between 25% and 250% over existing algorithms such as calendar queues. Taken together, these techniques provide an entity state generation mechanism capable of overcoming the challenges of Embedded Simulation in harsh mobile communications environments with restricted bandwidth, increased message latency, and extended message drop-outs.