The concept of rimware for the integration of IoT (Internet of Things) and the cloud to form a very powerful cyber-physical system is introduced. The specific implementation, called BlueRim, can scale from isolated NoTs to the global IoT. The rimware provides a middleware layer that enables gateway devices to establish connections between the devices and the cloud with roaming support while requiring them to prove their trustworthiness. The rimware has components that run on both the gateway and the cloud for each application domain. Rimware components are adapters running on the gateway, knowledge base running in the cloud; and access controller, also running in the cloud. As the gateway acts as the conduit, the purpose of authentication is to check whether the gateway is a trusted one. Access control is built on top of authentication, which is initiated from the device side. Rimware uses BLE profiles extensively as a fundamental mechanism for access control.

Modern scientific datasets present numerous data management and analysis challenges. State-of-the- art index and query technologies such as FastBit can significantly improve accesses to these datasets by augmenting the user data with indexes and other secondary information. However, a challenge is that the indexes assume the relational data model but the scientific data generally follows the array data model. To match the two data models, we design a generic mapping mechanism and implement an efficient input and output interface for reading and writing the data and their corresponding indexes. To take advantage of the emerging many-core architectures, we also develop a parallel strategy for indexing using threading technology. This approach complements our on-going MPI-based parallelization efforts. We demonstrate the flexibility of our software by applying it to two of the most commonly used scientific data formats, HDF5 and NetCDF. We present two case studies using data from a particle accelerator model and a global climate model. We also conducted a detailed performance study using these scientific datasets. The results show that FastQuery speeds up the query time by a factor of 2.5x to 50x, and it reduces the indexing time by a factor of 16 on 24 cores.

Modern scientific datasets present numerous data management and analysis challenges. State-of-the-art index and query technologies are critical for facilitating interactive exploration of large datasets, but numerous challenges remain in terms of designing a system for process- ing general scientific datasets. The system needs to be able to run on distributed multi-core platforms, efficiently utilize underlying I/O infrastructure, and scale to massive datasets. We present FastQuery, a novel software framework that address these challenges. FastQuery utilizes a state-of-the-art index and query technology (FastBit) and is designed to process mas- sive datasets on modern supercomputing platforms. We apply FastQuery to processing of a massive 50TB dataset generated by a large scale accelerator modeling code. We demonstrate the scalability of the tool to 11,520 cores. Motivated by the scientific need to search for inter- esting particles in this dataset, we use our framework to reduce search time from hours to tens of seconds.

This dissertation consists of three unrelated papers in labor economics. The first chapter documents the role of norms, both cultural and religious, in the fertility decisions of second-generation women in the US. Using two cohorts of immigrants (1970 and 2000s), I find that fertility declines among second-generation immigrants in the US are highly correlated with contemporaneous falls in total fertility rates (TFR) in Europe, implying that changes in the origin countries after parental emigration are still mirrored among current immigrants. This cross- country correlation is stronger for women from predominantly Catholic countries, which is consistent with immigrants from Catholic Europe sharing the Church's pro- natalist theology. The second chapter estimates the extent to which factor bias within manufacturing affects productivity growth across countries in the last two decades of the 20th Century. Skill-biased technological change (SBTC) implies that countries with more skilled labor and capital experience higher growth in total factor productivity (TFP), which is the case in both developed and developing countries in the 1980s. Labor-biased technological change is especially strong among the "newly industrializing countries" in the 1990s. These results are consistent with the empirical literature on skill-biased technological change, and may explain why "conditional convergence" of per capita income across countries is so slow. The final chapter examines the violence-reducing effect of development spending in Afghanistan. Using data from three distinct reconstruction programs and military records of insurgent-initiated events, the analysis finds that overall spending has no clear effects on the frequency of rebel attacks. Moreover, the types of development program most effective at reducing violence in Iraq -small CERP projects--does not appear to do so in Afghanistan

As Internet traffic continues to grow unabated at an exponential rate, it is unclear whether or not the existing packet routing network architecture based on electronic routers will continue to scale at the necessary pace. On the other hand, optical fiber and switching elements have demonstrated an abundance of capacity that appears to be unmatched by electronic routers. In particular, the simplicity of circuit switching makes it well-suited for optical implementations. Therefore, given the rapidly increasing traffic and optical transport capabilities, and the growing disparity between optical capabilities and Moore's Law, we would like to bridge this gap for future optical networks. In this thesis, we present a new approach to optical networking based on a paradigm of "coarse optical circuit switching by default" and "adaptive rerouting over circuits with spare capacity". We consider the provisioning of long-duration quasi-static optical circuits between edge routers at the boundary of the network to carry the traffic by default. When the provisioned circuit is inadequate, excess traffic demand is rerouted through circuits with spare capacity. In particular, by adaptively load-balancing across circuits with spare capacity, excess traffic is routed on their final destinations without the need to create circuits "on-the-fly". We call this new network architecture COPLAR, which stands for "[C]oarse [OP]tica[L] circuit switching with [A]daptive [R]erouting". Specifically, we first formulate the problem of circuit provisioning as a multipath utility max-min bandwidth allocation problem, which considers routing as an optimization parameter rather than input. The optimal and local algorithms proposed in the thesis are the first solutions to this problem. For traffic that cannot be handled by the default provisioned circuits, we apply an adaptive routing algorithm based on game theory to adaptively reroute the excess traffic over circuits with spare capacities. To evaluate COPLAR, we conducted extensive experiments using two separate real large-ISP PoP (point of presence)-level topologies, Abilene and GEANT. The results show that coplar has the ability to improve network throughput while significantly reducing electronic router overhead and the number of O/E/O conversions in the network. Finally, we show that our utility max-min bandwidth allocation algorithm can be extended to other network problems. In particular, we present a network security application called PSP (Proactive Surge Protection) that provides a defense against bandwidth-based distributed denial-of-service attacks

[No Abstract]