Jump to the main text
The Network Systems Laboratory (NSL) is part of the Hitachi’s Global Center for Social Innovation (CSI), located in Santa Clara, CA. NSL’s mission is to develop smart connectivity and edge analytics solutions for Hitachi’s IoT platform to accelerate convergence of IT and OT across various market segments such as Energy, Transportation, Manufacturing, Healthcare, and Smart Cities. We are currently developing technologies to enable an end-to-end secure, reliable and scalable IoT platform such as turnkey connectivity from sensors to analytics platform, cloud managed domain-dependent edge analytics solutions, fault-tolerant and optimized field networks for agriculture and renewable energy. We work closely with product and business units for developing PoCs through customer engagement and contribute to Hitachi’s patent portfolio in the process. Additionally, we also monitor and participate in various standardization activities including Industrial Internet Consortium (IIC), ISA-99 and 3GPP as well as promote ties with academic institutions for research collaboration. A sample of the most recent projects in NSL is,
Previously, NSL had participated in 3GPP cellular standardization discussions and meetings, focusing on RAN1, RAN2, and SA2. NSL has also been involved in IEEE 802.11 standardization activities, and played an active role in the formation of IEEE standard 802.11aa. NSL had also worked on enabling reliable video transfer over wireless systems and organized a demonstration at global Consumer Electronics Show (CES) in 2008 under the umbrella of Hitachi, Ltd.
The lab comprises of a dedicated team of researchers located in the Santa Clara campus for the Global Center for Social Innovation, North America (CSI-NA) and works closely with other researcher labs in CSI-NA, and Hitachi’s Center for Technology Innovation (CTI), in Tokyo. NSL also promotes ties with leading academic institutions for research collaboration such as Stanford University and Carnegie Mellon University.
An IoT platform that enables seamless connectivity from sensors to the analytics cloud is critical for enabling IT and OT convergence and optimization. Typical functions of an IoT platform include adaptive filtering of sensor data at the sensor nodes themselves or at the edge gateway, cleansing and aggregation of sensor data and distributed decision making at the edge gateways for minimizing latency of critical applications. In addition, each vertical application requires specific features for data filtering and edge analytics depending upon volume of sensor data, cost constraints for transporting data to the cloud, security considerations and need and feasibility of implementing a distributed decision making architeture. To facilitate the seamless integration of industrial IoT infrastructure and devices to the platform, NSL is developing specific edge computing solutions that cater to various verticals such as connected fleets, micro-grids, wind turbine, and agriculture. NSL is also developing a prototype IoT platform that can be deployed at customer site for validation of proposed solutions. The platform includes a wide variety of sensors both commercial and custom made that sense different kinds of physical phenomenon such as temperature, humidity, vibrations, light etc. and the deployment can be scaled over a large enterprise area. The sensors communicate to a tiered system of gateways which supports “turnkey connectivity” for a variety of wireless protocols such as WiFi, ZWave, Zigbee and so on. Additionally, the proposed platform allows the end user to develop their own applications to customize the platform for their specific needs.
Many governments in Europe and US have mandated some form of connected driving. NSL is looking into multiple technical aspects of connected driving including managing efficient connectivity and edge filtering for fleet convoys and cargos. NSL is developing a platform to efficiently collect and process data from a large number of collected vehicles over the cellular network and perform in vehicle edge analytics of this data to extract meaningful information.
With millions of devices connecting our homes, cities, factories and infrastructure to the cloud, security becomes a major concern. The problem is compounded for industrial IoT scenarios where legacy OT systems are increasingly being taken out of their silos and connected to internet based modern IT systems. It is not clear if the traditional models of IT security will suffice for securing IoT devices and applications. NSL is leading a fundamental rethink on IoT security for all layers from the sensors to the gateways and the cloud. NSL believes that security should be an integral part of the IoT system design and not a later add-on feature. NSL is also collaborating with CyLab, Carnegie Mellon University for this project.
Network operators and vendors are struggling to find ways to meet the exploding demand for data and multimedia content, triggered by the proliferation of tablets and smart phones. The ever increasing demand for multimedia services creates traffic congestion, and reduction in the quality of service, not only at the RAN side, but also at the CN and PDN sides. NSL proposes highly granular, scalable and predictive traffic management solutions for network operators to deal with congestion. NSL’s algorithms consider the telecom networks in their environments, and take into account data from external factors such as road congestion and weather.
NSL’s algorithms further propose pre-emptive solutions at the RAN and CN to optimize the network utilization. NSL has developed demos and proof of concepts for RAN-aware network optimization. NSL members generated a number of invention disclosures on the topic and showcased their solutions in a recent demo in MWC 2015.
Enterprise WiFi networks are managed networks that are experiencing saturated performance due to the increasing data traffic and the potential offloading from cellular networks. Mechanisms to efficiently manage the network include load balancing among APs, interference management through channel allocation and power control, as well as optimization of the CSMA threshold to increase the number of concurrent transmissions. Enterprise WiFi networks still suffer, however, from increased congestion, and interrupted services and advanced techniques are needed to analyze the WiFi networks and optimize the delivery of different types of traffic with an acceptable quality of experience. NSL has generated a number of patents on this topic and developed proof of concepts for real-time network monitoring and optimization solutions including robust delivery of video streaming content.