Network centrics

Network Centrics is the emerging discipline of applying and enhancing networking and computing concepts and implementation technologies in a variety of social, business, content distribution, and communication domains involving interactions amongst the domain entities. Network Centrics acts as a catalytic agent resulting in significant improvements in the definition, modeling, comprehension, performance, and effectivity associated with these interactions. This topic has gained importance because of increasing connectivity at individual, corporate, and government levels. Network Centrics is supported by a sustained growth of Internet in conjunction with continued advances in efficient computing algorithms, cost-effective data-persistence solutions, and high-performance data network options that can now be leveraged in a variety of domains.

Overview

Network Centrics focuses on the central role of integrated networking and computing as a premise, facilitator, and accelerator for a rapid, sustained, cost-effective and managed improvement of interactions among the various entities constituting a domain. It formalizes, automates, and extends the fundamental concept of network effect that describe the underlying connectivity among large number of entities in a wide variety of fields ranging from social groups, electrical transmission systems, spread of pandemics, metabolic pathways, gene expression, evolutionary behavior, and various Internet based large applications. Closely associated with such fundamental network connectivity is the popular concept of six degrees of separation (that all human beings are on the average only six acquaintances away from any other human being in the world) that is now part of common vernacular.

Improving Interactions among the Domain Entities

Underlying connectivity among the entities which are spatially or temporally separated from other entities is severely inefficient, thus limiting the network effect for the associated interactions. The network effect can be improved by using an automated information infrastructure as a substratum for bridging geographical and temporal distances in a fast, reliable, and secure fashion. Networking infrastructures have evolved from smoke and light signals, telegraphs, telephones, mobile telephony, to Internet applications leveraging key advances in electro-mechanical transducers, semiconductor, magnetic data storage, software programmed computers, wireless communication, and algorithms. Although technologies such as computer hardware and software have been critical in the growth of digital data and telephony networks, the two broad fields of networking and computing have evolved independently, thus limiting their potential in certain domains. Network Centrics consolidates the principles, tools, and technologies of the networking and computing fields for a unified approach towards improving domain interactions.

The meteoric rise of social networking web sites such as My Space and Facebook has demonstrated the power of networking (fast connections for large number of users) and computing technologies (databases, search algorithms, matching algorithms, user interface). Such an infrastructure serves as a potent catalyst which can improve connectivity and improve associated interactions with several orders of magnitude. Another application of the Network Centrics concepts is exemplified by a mature and scalable trading platform provided by Internet auction sites such as eBay that efficiently interconnects the buyers and sellers across geographical and temporal boundaries. This Network Centric platform requires efficient implementation of auction algorithms, databases for storing goods and current bids and email software for automatic notifications used to improve the auction process all working over a large area data network. An extensive use of both computing and networking technologies enables a reliable, transparent, secure and well defined way of millions of interactions on a daily basis for a large number of participants interested in finding and buying specific physical goods on the one hand and sellers on the other who have profit making objectives. This is in stark contrast with the local bazaars of yesteryears with at most hundreds of people participating in handful of auctions in a session.

Besides these recent Internet phenomena, now part of the day-to-day life in all continents and most countries, there is also a profound change in the role of networking for national defense. Many large countries, led by United States, have unequivocally stipulated that the current defense strategy should be radically transformed with information networking being the central substratum for command, control, intelligence, sensing, and reconnaissance. The scope of such network-centric warfare goes far beyond policy, directives, training, and logistics to actual remote real-time sensing and shooting. This involves a robust, secure and comprehensive integration among all sensor integration, data collection, data storage, information analysis, decision making, and weapon system activation tasks across large geographical distances and multiple services and coalition partners. In fact, the term Network Centricity itself has emerged from the set of many policy documents, starting with Adm. William Owens introduction of the term “system of systems” followed by Network Centric Warfare as discussed by Vice Admiral Arthur Cebrowski and John Gartska, and detailed in the book “Network Centric Warfare”). Netcentric approach is a now a key facet of the strategic transformation of US defense forces with information superiority.

Extending Basic Networking Concepts

Network Centrics extends the fundamental concept of domain networking that exists, though in a constrained fashion, independent of any automation provided by information (computing and networking) infrastructure. This extension is initiated by adding specific networking concepts such as addressing, routing, multicast, persistence, robustness,^{[disambiguation needed ]} and assurance. This is further enriched by leveraging computing concepts such as discovery, mediation and search, collaboration, filtering and data mining algorithms with associated automated implementations using computer and software technologies. Adequate sensor and actuator components are used to interface the domain entities with the underlying automation infrastructure. The use of service level agreements for interactions ensures quality of service, for a specific price, and associates information assurance for privacy, authenticity, and non-repudiation. These technical concepts are well understood and widely implemented with a large number of technologies as epitomized by existing information automation infrastructures (networks). This includes the Internet with TCP/IP and HTTP family of protocols and the mobile communications devices based on GSM and CDMA standards providing voice, email, paging, and other applications. The implementation of these infrastructural networks has been possible with extensive research and development over the past four decades in packet networking, mobile cellular architectures, computing algorithms, point and click user interfaces, data processing techniques and Internet web applications. Though landline phone networks have been around for more than a hundred years and at least initially played a major role in accessing Internet with dialup connections, their importance is diminishing rapidly because of their inherent efficiency, cost, and flexibility issues.

Studying and organizing the interactions among domain entities can be facilitated by judiciously applying one or more networking concepts such as queuing theory and computer simulation techniques. For any performance improvement, however, automated technologies such as wide area communication, algorithmic manipulation, and persistent storage of data in transit, bring us from basic and simple networking to the field of Network Centrics. The use of appropriate sensor technologies which can transform many chemical, biological, physical, and situational data into electrical signals facilitate linking with data collection networks for further information aggregation, processing, knowledge mining, and decision making. Today it is mostly the computer and network technologies manifested in well known products such as computer servers, PC, web browsers, database management system, email, paging, data mining tools, TCP/IP protocols, routers, optical communication links, mobile PDA, etc. that are deployed for a network centric performance boosting for interactions occurring over large distances and time scales. Besides these widely deployed computing and data network technologies, there are other emergent technologies that are also relevant for Network Centric improvement of domain interactions.

Emergent Technologies and Domains

For certain fields, the spatial and temporal scales are much shorter compared to our day-to-day networking scenarios using Internet and mobile phones. An important example of such network centric domain is associated with chemical, biological and physiological interactions among the molecules, cells, tissues, and organs within the human body. It is already possible to place electronic devices such as pacemakers, and insulin dispensing pumps that automate interactions within the human body with suitable sensors and actuators, data collection wired network, and micro-computer based processing without any external support. The development of advanced microscopic devices using nano technology for electrical, mechanical, and chemical transduction will revolutionize human health care for disease prevention, monitoring, and treatment. Such futuristic devices will involve interactions at molecular and cellular levels occurring at sub-second time scales. Other healthcare interactions such as disease biomarkers monitoring, patient-physician encounters, adverse drug reactions and clinical trials for drug discovery are also witnessing a ground-breaking revolution supported by network centric improvements. This is best epitomized by the national initiative to interconnect healthcare entities such as medical practices, hospitals, laboratories in federated networks called National Health Information Networks (NHIN).

Network Layering and Protocols

The fundamental primitives of a Network Centric domain comprise the entities within the domain and the interactions that occur across these entities. At a more detailed level, the various entities can be partitioned into multiple categories with respect to their roles, responsibility, and interests. Any complex domain is also likely to display network centricity at various levels or layers where an interaction can be further decomposed into more detailed sub-entities and/or sub-interactions at a lower layer. This layering is one of the most basic principles of data networking.

The data networking interactions are typically associated with specific protocols that give structure to the way information is organized, sequenced, and exchanged. A well known networking protocol is Internet Protocol (IP) that is supported by every Internet networking node at the network layer of the Open Systems Interconnection (OSI) model. The OSI model prescribes a seven layer model for data networking. The physical layer defines the electrical and optical behavior between two nodes directly connected. Any error detection and recovery along with suitable data packet framing is covered in data link layer. The network layer covers direct and indirect connectivity between any two nodes with associated addressing and routing capabilities. The transport layer provides reliable data communication across two end nodes, typically called hosts. The upper three layers, session, presentation application, involve end user computer applications.

Protocols are not just for data network implementation. As an example, a clinical protocol could also be used in a clinical trial that defines how patients are selected, medical experiments defined, drug dosage administered, adverse reaction monitored, clinical data collected, and results processed for discovering new drugs for treating a disease. It is myopic, however, to see the automation of drug development interactions as just a computing problem (and supported by a centralized automated information system). This may be one of the reasons why clinical trials continue to be expensive despite the availability of clinical trial management software. Such protocols or interactions involve large number of patients distributed over multiple geographic areas working with their respective physicians, under the scrutiny of regulatory bodies at local, regional, and national levels, and finally sponsored by a profit-oriented pharmaceutical firm. This creates a complex domain that requires networking and computing technologies in a unified fashion for higher performance and improved information security for all interactions.

Data Network Concepts and Technologies

Addressing Routing Quality of Service Service Level Agreement Multicast Management Policy

Computing Concept and Technologies

Server Web Browser Email Protocol Web Protocol Store and Forward Persistence

Information Assurance

Authenticity Privacy Non Repudiation Robustness

Emerging Technologies

Adaptive Network Implementations Nano Technology Web 2.0 Semantic Web Ontology Network Centric Platform

Network Centric Domains

Social Networks Goods Exchange Health Care (Disease Prevention, Detection, Treatment) Medical Research (Proteomics, Genomics) Drug Discovery (Clinical Trials) Education Warfare Home Land Security

Organizations Involved in Network Centrics

Computer Companies Networking Companies Health Research Institutes Department of Defense

References

Barabasi Alberto-Laszlo, “Linked, The New Science of Networks, How Everything is Connected to Everything Else and What it Means for Science, Business and Everybody Life,” Pursues Publishing, 2002, ISBN 0-7382-0667-9.