SO96.3.16

ABSTRACT

The future brings a continuing need for state-of-the-art cost effective operational systems. These systems must support and keep pace with escalating Satellite and Telecommunications (SAT) services. Space programs continuing to develop futuristic computer and telecommunications system will also demand advanced operational systems. To be effective operational system designers must centre on using rapid prototyping, collaborative computing technologies, and Commercial Off-The-Shelf (COTS) products that operate on open platforms. This paper describes a New Century Architecture (NCA) representing a typical future operational system using where possible vendor independent, shareware and COTS systems. The time frame between 1996 - 2010 fits an NCA development cycle for two reasons: (1) it permits existing systems to upgrade and not be encumbered by transition problems during current day-to-day operations, and (2), a minimum of 3 to 4 years is needed to bring NCA innovation cycles from initial concepts to general availability within operational applications.

The Internet causing dramatic changes in the business world is fundamentally altering the organisational structures of our institutions and the living conditions of our society. Looking toward the next century, it will be imperative for private enterprises, telecommunications organisations, the space industry and countries to exchange technology on world wide basis. Business will lead demand for (SAT) services and international data communications will remain the fastest growing parts of business traffic.

SAT Demand will force a significant shift towards end-to-end cost effectiveness following similar digital transmission and storage of information trends witnessed in telecommunications. Continued deployment of digital technology permitting different modes of information (voice, numbers, text, image, graphics, video) to be transmitted and manipulated via hardware and software components will result in lower development and operating costs using less physical space and fewer people.

Developers will rethink and re-engineer basic SAT processes with the goal of reducing overall programme expenditures. In all respects this may become a necessity because of declining government budgets, which as everybody knows has already resulted in business mergers, reconsideration, reduction or complete cancellation of various projects.

Five business "Factors" will shape SAT development

Corporate Networking - Common user networks coupled to the Internet will support new business applications; a growing emphasis will be placed on using SAT networks to implement intranets for increasing access to corporate databases, building links to customers, suppliers and so on.

Satellite Backbones - Fixed ground and mobile user stations provide economic justification for combining dispersed networks to support corporate resources. INTELSAT, for example has already demonstrated that its fleet of 24 satellites are able to provide global access to the Internet.

Privatisation - More and more corporate and government organisations will take responsibility for their own SAT networks; installing satellite links, high speed multiplexors, private transmission and specialised switches saves time, money and increases control of information.

The PC revolution - Computing and SAT becomes down sized and distributed. The Internet interconnected to the PC, whether at home or in the office opens a inventive communications era requiring new SAT structures to overcome limitations on existing networks. WEB subscribers has gone from 2 million in 1994 to more then 30 million by June 1996. Introduction of a WEB PC may further advance the revolution by offering a cheap computer that discards heavy memory use, OSes, specific applications, and bloated platforms.

Data communications growth - In 1985 corporations were spending 30% of their telcoms budgets on voice and 20% on data. Voice has been growing at six percent per year and data at 40% percent. It should also be noted that GSM is growing in a similar manner and it is estimated that by the year 2000 half the information over the GSM network could be data.

The simplest way to describe the growth in SAT systems is by one word "access". Access to SAT networks has created three businesses: (1) the transmission business, (2) the storage business and (3) the understanding business.

By the year 2005, almost all segments of society will include individuals who will be in the understanding business. These individuals will be the users of SAT systems - and be affected by or be dependent upon SAT systems in their daily work or leisure time. Requirements will vary greatly, depending on their jobs, and how interactive they become. SAT users and their modi operandi can be described by large number of characteristics, which this paper will not go into. These characteristics are not by any means, independent of one another, but many of them are nearly so. Even with tremendous change in technology the following facts will remain:

• users' needs are not adequately fulfilled today; and
• users' needs and expectations will increase and expand with time.
  

Tomorrow's SAT systems must adequately serve its users. It must provide functions in a manner users expect, when needed, and at a cost users consider reasonable.

SAT based information systems including sensor monitoring, data collection, and analysis already has produced major impacts upon operating environments. SAT driven computer models of the weather system are greatly improving long-range and large scale weather prediction. Imaging of air, ground and water pollution provides a better understanding towards forecasting. These missions include defining and measuring air quality and the effects of pollutants, thereby, introducing methods to control, for example, motor vehicle traffic patterns, stationary pollutant sources, and so on. Operational aspects of these systems requires sorting large data files in a highly structured fashion while providing their manipulation in a user-oriented language, plus accurate handling of computational problems, rapid turn around time, and time sharing of hardware and software resources.

Developing global interfaces to accommodate this constantly growing demand of unpredictable traffic volume is a primary NCA design goal. Designs must cope with increased intermingling of different traffic types, some of which are not even available for study, coupled with Internet traffic, valued added services, data bases, applications software, etc. How will NCA systems be designed, measured, dimensioned, controlled? Present lack of traffic level and characteristic data makes design, forecasting and control problems very difficult indeed. NCA operational structures must be insensitive to traffic characteristics and provide rapid and convenient rearrangement and reprocessing facilities to maintain performance standards during periods of growth or change.

NCA development must have a coherence between Design, Development and Operational Phases. Employing NCA technology, tools and methods provides an opportunity to bring about a condensed life cycle resulting in reduced cost and time without sacrificing quality.

Achieving this objective requires NCA building blocks to be small dedicated modules, each designed with processing and performance levels to accomplish a single task. Each module contains a high degree of regularity among functions implying an ability to share logic among other modules, thereby, reducing both maintenance and operational manpower costs whole instilling a strong effect on ease of programming.

Developing small dedicated modules can decrease elapsed project time by between 30% to 50%. Total effort ( i.e. number of man hours) for the same development process can be expected to decrease by between 25 % to 40%. Quality is greatly enhanced not only because of size but also due to simulations performed within all development stages of an NCA life cycle.

Significant reuse of hardware and software components are incorporated in NCA modules so that functions can be shared. Cost effectiveness is achieved by using low-cost COTS components coupled with an ability to install additional capacity, when required, in small increments.

Simulations becoming standard NCA design features provide methods towards switching from traditional programming life cycles to systems employing System Description Languages (SDLs). SDL changes programming concepts and provides for increased valuation during early design stages, a must for NCA systems.

A simple overview of SDL compared to a traditional system design is shown in Figure 1

Figure 1. SDL System Structure

As shown in Figure 1, SDL development proceeds along through 5 phases.

• In the systems analysis stage, designers analyse requirements and the interactive behaviour of a NCA system and build user interactions in chart form.
• System Design stages build Specifications which involve building an SDL diagram of the blocks and processes that describe the system.
• The simulation stage builds interactive models of the systems operation so to test design assumptions.
• The verification stage consists of internal SDL logic checking the SDL diagram for syntactic and semantic errors, which can be corrected interactively.
• Once the system has passed validation, it can be instructed to generate "C" code or a specialised simulation language which can be used as input to a simulations model, which will test and reveal run time logic errors that escaped the validation phase.
  

Results are C-coded test suites that are independent of both the target system and the application. This means that the generated code suites any test structure supporting "C". It is estimated that SDL Real time development tools can improve productivity in the order of 50% to 60 %.

Building a SAT mission can be accomplished using a set of integrated COTS tools. To illustrate the concept, we will use three separate COTS tools supplied by Analytical Graphics, Inc, Satellite Tool Kit (STK), Satellite Tool Kit Programmers Library (STK/PL), and Satellite Tool Kit Visualisation Option (STK/VO). Figure 2 shows the relationship between these three modules.

Figure 2. COTS Tools

The base STK system lets the user:

• Propagate vehicles ( satellites, aircraft, missiles) to determine position and attitude data
• Display coverage areas
• Calculate and display access times between defined systems components

In general, the user can generate paths for vehicles (both orbiting and non orbiting) to determine access conditions between vehicles, targets, and facilities. Figure 3 shows the possible options that a user can select, for example, using a vehicle, i.e., movable land, sea, air or space objects

Figure 3. Vehicle Objects

The interconnection of STK to STK/VO provides the user with a three dimensional viewing capability that provides mission and orbit analysts an intuitive view of complex SAT mission and orbit geometry by displaying realistic 3D views of space craft, sensor projections and orbit trajectories. Interconnection of STK/PL provides the user a set of tools that contain high level Astrodynamics, Graphics and User Interface routines and low level functions such as list and stack management, database and parsing routines.

It is the purpose of the software structure to provide a uniform framework for developing a SAT mission scenario within a heterogeneous computing, communications and applications environment. The communications environment includes several different individual network designs.

Interconnection to user applications can be accomplished by adding an Inter process Communications Module (IPC). IPC enables a user to work with STK in a client-server environment. Through IPC, a user application can load a vehicle into a STK scenario, determine access intervals between objects, and return those intervals to the user application for specialised analysis and processing. Real-time information, such as telemetry data from an actual vehicle can be passed to STK to build a scenario, in real time, complete with attitude and position information.

At the workstation level, IPC can be connected to either a Unix or TCP/IP Socket, while at the PC level, the interconnection can be Ethernet or Token Ring. The approach taken is to integrate STK tools into a resource sharing computer network under a single monitor and file system and make all

STK tools uniformly accessible to designers, programmers, project managers, and operational personnel. The communications structure sets the states for the interconnection and transmission of data and a status indicates (1) what data has been successfully sent, (2) indication when data can be sent, and (3) indicate what data has been correctly received, and which sends or receives may be in trouble and the nature of the problem.

This configuration is shown in Figure 4.

Figure 4. Communications Interface

To summarise, COTS systems coupled with technological advances in storage/processing logic and interconnecting structures used in an NCA system will reduce systems costs dramatically. NCA using COTS will attract a large community of users. Also users of existing systems will see the benefits of adding functions by installing COTS systems. The next section demonstrates how the COTS system explained above fits into an NCA system

Major design goals for the NCA are high processing power, large memory capacity, high reliability, low cost, modular structure and flexibility. These goals can be realised through a structure where each module has its own operational structure and can perform tasks independently using existing COTS software wherever possible. Developed code would be in the form of object-oriented structures to provide flexibility and reuse among, for example Telemetry, Tracking Control and Technical Operations Control.

Each Module by means of a Network Interface has full access to other modules providing exchange of control and data information. The structure envisioned would be Windows NT operating with a PC base connected to a network. The PC interconnected through the network to an SQL Server provides a database structure providing information links between, for example, operational analysts and mission management. TCP/IP, Ethernet Routing and MAC OS support are all part of the standard NT package. Dual network interconnections can be provided to achieve desired access redundancy goals, with internal logging and backup functions for all information that is critical.

Figure 5 shows a block diagram of a typical architecture.

Figure 5. NCA Architecture

Internally, NCA is a multiprocessing system, but it is a unique application of multiprocessing in two respects. First, operational personal are totally unaware of its operating nature. Second, each Processor Module (PM) is dedicated to a specific computing function. For example, within the Master Control Unit, one process is dedicated to telemetry processes, another to command encode, another to command decode, whereas PMs connected to the network operate as single units with one processor dedicated to satellite control, another contains required analysis tools, another providing administrative functions, and so on. STK, for example, could be used to analyse changes in orbit positions received from telemetry data against original models calculated for a mission.

All PMs have a queuing mechanism for receipt and transmission of messages, and all PMs can be active simultaneously. Data passes from PM to PM as different activies occur. The Front End Comms Interface (FECI) acting as a fixed station is the intermediary between gateways and SAT communication, e.g., sensor data, data communications, control information utilising down link, up link or terrestrial transmission. Currently, many of these front end systems are proprietary stand alone structures but they will be replaced by sets of logic cards being inserted into PC expansion slots running under Windows NT. NT Systems equipped with Alpha or MIPs CPUs or even multiple Pentiums can overcome throughput requirements offering a price/performance advantage.

This paper has portrayed changes in operational structures as a result of a shift to COTS software. Systems generated by individuals and organisations creating and taking advantage of the opportunities provided by COTS software structures will become sufficiently large and far-reaching to collectively comprise a technical advance to operations planning.

User needs for operational flexibility, allowing continuing adjustment to exiting systems will continue to grow. An important contribution to this process is the technical and cost benefits of using COTS. The next century will see more and more operational services provided by new generations of COTS . These will emerge from new service provider organisations, who are willing and able to put technology to work to satisfy a growing number of users. The fundamental point about using COTS is that there are many technical and cost reduction opportunities presented by the integration of this technology. Whether or not "COTS" is good for you, depends on how you use it.