6. Standards Technology Forecast

This section describes and discusses a variety of emerging technologies expected to result in new standards that will have a significant effect on the USIGS. This section is not intended to be a comprehensive analysis of technology trends in general.

The first part of this section is a synopsis of the current NIMA technology forecast. This is followed by discussion of these technologies:

• Distributed Computing
• Activities at the OGC in Open GIS technology
• Collaborative Computing
• Security in a distributed multi-domain environment

Each of these will be discussed relative to their status at the time that this document was completed, expectations for the near future, and their specific effects on the USIGS.

6.1 Synopsis of NIMA Imagery and Geospatial Technology Forecast

The current NIMA technology forecast, published in June 1997, is called "R&D Strategic Thrusts: FY96 Imagery & Geospatial Technology Baseline" [NIMA97]. The forecast is organized around four functional areas that are central to NIMA and USIGS:

1. Exploitation and analysis technologies
2. Information and data handling
3. Implementation tools and methodologies
4. Collection technologies

For each of these areas, a forecast is presented in three time frames: until 2002, 2002-2008, and 2009-2014. Table 6-1, Table 6-2 and Table 6-3 show the forecast through 2002.

Technology Sub-Area	By Target 1 (2002)
Automated Image Examination	Limited ATD/ATC on workstations Automated 3-D wireframes Limited 4-D video exploitation
Softcopy Exploitation	Digital capability Integrated video exploitation Historical, comparative & predictive decision support tools
Data Fusion and Analysis	Manual/semiautomated visual fusion
Information Visualization	Fly-through capability (Virtual Reality display)
Spectral Phenomenology	Limited Spectral understanding and application
Image Display and Reconstruction	Advanced 2-D/aided 3-D displays
Multimedia Products	Semi-automated product generation
Collaborative Exploitation and Analysis	Peer-to-peer and team collaboration

Technology Sub-Area	By Target 1 (2002)
Integrated Imagery/Geospatial System Management	Multiple (2+) management systems Integrated softcopy exploitation management Automated multiple objective/multiple system collection nomination Network-based imagery/geospatial product dissemination for EAC (echelons above Corps) tactical users Smart compression techniques for product dissemination Customized sensor-to-shooter packages
Mass Storage and Management	Operational 3-dimensional storage technologies Ruggedized high capacity storage systems (~petabyte) Context-sensitive archiving State-of-the-art optical disk storage Partial regional imagery and geospatial product archives
Information Management, Discovery, and Retrieval	Context-based search and retrieval Text and limited imagery/geospatial data NIDR Intelligent SW agents Limited-domain BOK for imagery/geospatial and collateral data
Communications Equipment and Networks	High Capacity Fiber data connectivity > 2.5 Gbps in major world population centers High capacity mobile (radio/VSAT/direct broadcast) to deployed units at 10-45 Mbps Implementation of imagery/geospatial-capable LANs (> 150 Mbps) Initial implementation of network-based imagery and geospatial product dissemination system, including tactical users
Communications Security	Dedicated, system-high networks Prototype operational MLS systems Advanced firewall technologies High-capacity (>625 Mbps) key-agile encryption schema Use of session encryption software for time-sensitive data
Data Compression and Decompression	Wide application of state-of-the-art commercial compression techniques to literal/SAR imagery, video, voice and geospatial data Imagery/geospatial data compression ASICs Compression supporting successive refinement of imagery/geospatial data
Computer Hardware and Software	Latency-tolerant, distributed operating systems Object-oriented or transaction-based operation Low-latency parallel architectures (>100 GFLOPS throughput) Prototype optical computing in selected applications Advanced 2-D graphical user interfaces User-tailorable SW algorithms

Technology Sub-Area	By Target 1 (2002)
Software Engineering Tools	Advanced Computer Aided Software Engineering (CASE) Tools providing platform independent programming Transparent applications with CORBA compliance Integrated testing environments Fuzzy logic programming Distributed, collaborative engineering environments
Simulation Tools and Methodologies	Development of a high-fidelity model of the exploitation process Analytical simulation tools for imagery/geospatial development and acquisition Protocols and open architecture standards for distributed interactive simulation and interoperability Use of modeling and simulation to support requirements and procedures development Scaleable simulations
Training Tools and Methodologies	Incorporation of advanced learning technologies Fully embedded training systems Context sensitive/adaptive training Integration of imagery/geospatial products into user training systems Virtual reality training environments at the operator level

6.2 Distributed Computing

At the time of this writing there existed four predominant technologies for implementing distributed computing, one of which is oriented toward procedural distributed computing (but which supports distributed object computing (DOC)) and three that are oriented toward DOC. All are discussed in the following paragraphs.

6.2.1 Distributed Procedural Computing

Distributed procedural computing is predominantly accomplished using the Open Group-defined Distributed Computing Environment (DCE). It is important in the context of this discussion because DCE services are often used with an object-oriented "wrapper" as a quick way to provide CORBA services. In particular the Domain Name Service (DNS) and the Security Service (based on Kerberos) are often used by CORBA vendors to provide a quick solution, while taking more time to develop "pure CORBA" implementations. The DCE Transaction Service is also sometimes used in CORBA service implementations.

Also important to note is that the Microsoft Distributed Common Object Model (DCOM) protocol (see Section 6.2.2.1 below) is based on the DCE Remote Procedure Call (RPC) and that this same RPC is an allowable, but not recommended, wire protocol in CORBA implementations.

6.2.2 Distributed Object Computing (DOC)

6.2.2.1 Microsoft COM+

Microsoft ActiveX contains within it a wire protocol for distributed computing known as the Distributed Component Object Model (DCOM). This particular protocol is an extension of the DCE RFC that is intended to allow for distributed computing using the Common Object Model (COM) supported by Microsoft. This particular protocol is quite acceptable for a small closed-environment distributed computing capability, but has several serious liabilities when used in a Wide Area Network (WAN), particularly with respect to security. Thus, beyond limited use in a workgroup environment, it is not particularly suited to use within the USIGS. (NOTE: This does not mean that other elements of ActiveX should not be used.)

It is known, however, that Microsoft is working on a replacement for the DCOM that is expected to be available for use in 1999. It is too early, however, to speculate on its form and functions.

With respect to the DCOM effect on the standards environment, a specification for DCOM was submitted by Microsoft as an informational Request for Comment (RFC) to the Internet Engineering Task Force (IETF) in 1996. This type of RFC has a six-month lifetime, after which it is no longer available as an IETF document. Microsoft has allowed the RFC to expire and had not superseded it with any other specification by the time of this writing.

6.2.2.2 OMG Common Object Request Broker Architecture (CORBA)

The Object Management Group (OMG), a consortium of nearly 800 members (including Microsoft and Sun), has been developing the specifications for CORBA and its environment since 1989. The OMG has recently been approved as a Publicly Available Specification (PAS) submitter by the International Organization for Standardization (ISO)/International Electrotechnical Commission (IEC) Joint Technical Committee 1 (JTC1). This allows the OMG to submit specifications directly to JTC1 and its subcommittees, without going through any national body and without being edited by any subcommittee or working group of the JTC1. By being submitted as a PAS, a specification will take on a status equivalent to a Draft International Standard (DIS) and may be balloted by any JTC1 subcommittee. If such a specification completes balloting successfully, it will become an International Standard (IS).

The OMG is considering its options, but it appears that the first OMG specification to be submitted to the JTC1 will be the subset of the CORBA version 2.1 specification (OMG document formal/97-09-01: CORBA/IIOP 2.1 Specification) [CORBA97d] that defines the Internet Inter-Orb Protocol (IIOP). IIOP is the protocol that allows Object Request Brokers (ORBs) to interoperate.

IIOP was initially designed to provide interoperability between ORBs by passing object references between the ORBs. This enabled a client application to invoke objects on a server rather than on the client. The IIOP specification is being augmented with a "pass-by-value" capability that will allow for the transfer of objects between ORBs, alleviating several problems including easing the communications load associated with the transfer of collections of fine-grained objects. It is also expected that client-side garbage collection will be added to the IIOP specification in 1998. Other refinements can be expected over time, but it is unlikely that there will be other revisions in the near future.

A subset of the CORBA version 2.0 specification [CORBA97a] has already reached the IS status, the Interface Definition Language (IDL). In ISO, it is known as the Reference Model for Open Distributed Processing (RM-ODP) IDL (ISO/IEC IS 14750). RM-ODP IDL is syntactically and semantically identical to CORBA IDL, although the documents themselves differ slightly.

OMG also has a number of Task Forces addressing specific information domains including electronic commerce, finance, manufacturing, medicine, telecommunications, and others. These Task Forces will adopt domain specific specifications for the CORBA environment.

6.2.2.3 Remote Method Invocation (RMI)

The third prominent distributed object capability is the Java Remote Method Invocation (RMI). The Sun Microsystems JavaSoft Division has defined the JavaSoft-proprietary RMI wire protocol to enable the interoperability between Java objects (applets). This protocol is meant for Java-to-Java interoperability only and is both more and less than CORBA IIOP. It is more than IIOP in that it includes client-side garbage collection. It is less than IIOP, because it lacks some of the functionality of IIOP and does not support languages other than Java. JavaSoft has stated an intention to continue the development of RMI to provide functionality found in IIOP and not in RMI, while OMG will be enhancing IIOP to include client-side garbage collection. JavaSoft will then include both in the Java Development Kit (JDK), with RMI to be used for Java-to-Java interoperability and IIOP to be used for Java-to-everything-else interoperability.

6.3 Open GIS Consortium Activities

A Geospatial Object Model is being developed by members of the Open GIS Consortium (OGC) who are from both the user and vendor community, in conjunction with developers of the USIGS architecture. The model, which will be documented in a technical report, is focused on services that maintain, provide access, manipulate and exploit MCG&I data (e.g., imagery and feature data). This modeling effort provides a forum where parties involved in developing portions of the USIGS Architecture can discuss and collaborate on what software components are needed to support various functions of USIGS. The model being created is not attempting to provide implementation level detail. That is left to profiles being developed under the auspices of the UIP effort. The object model should be considered as guidance only at this time and for the foreseeable future.

The object model has been constructed with the goal of facilitating collaboration among the various NIMA Development Programs. Collaboration has been done with different groups implementing portions of the USIGS System Architecture. Various ongoing implementation efforts are developing a set of specifications with component interfaces defined in the Interface Definition Language (IDL). These efforts are producing object models to serve as guidance in generating the IDL. Portions of these various groups' object models have been examined and adopted as appropriate for this effort. The two that have been examined to date are the Geospatial and Imagery Access Services (GIAS) and the Geospatial and Imagery eXploitation Services (GIXS). The GIAS is being developed for the NIMA Libraries Program, while the GIXS is being developed for various groups of the NIMA Exploitation Tools Division. These object models are being harmonized with the model being generated as part of this effort.

Another major effort to be undertaken, but not yet initiated, is the harmonization of this modeling work with the development of the Conceptual Data Model being defined as part of the USIGS architecture.

Table 6-4 shows services that will be available for various UTA components by the end of 1999, and the specifications that are being developed by OGC or OMG. Table 6-5 shows other services that will be available in 2000 or later.

Applicable UTA Component	Applicable Specification	Status	Responsible Organization(s)	Comments
Shared Domain Components - Telecommunications Domain	Telcom Log Service	RFP has been released.	OMG	Adoption of this technology is expected in 1999.
Geospatial Domain Access Services	Gridded Coverages	RFP released	OGC	Adoption of this technology is expected in 1999.
Geospatial Coordinate Transformation Services	Geospatial Coordinate Transformation Services		OGC	OGC is in the process of developing a technology RFP.
Imagery Manipulation Services	Gridded Coverages	RFP released	OGC	Adoption of this technology is expected in 1999.
Imagery Manipulation Services			OGC	OGC is developing an abstract specification for geospatial and imagery portrayal services.
Imagery Exploitation Services			OGC	OGC is in the process of developing an abstract specification for Mensuration Services.
Geospatial Display Services			OGC	OGC is in the process of developing an abstract specification for this service.
Interoperable Name Service		Initial submissions received	OMG	Expected to replace the Naming Service in 1998 or 1999
Compound Presentation and Interchange Facility	CORBA Component Model	Pre-Initial submissions received	OMG	This facility is often referred to as "CORBA Beans."
Calendar Facility	Calendar Facility	RFP issued	OMG	Adoption of technology is expected in 1999.
Shared Domain Components -- Financial Domain	General Ledger Facility	RFP has been released.	OMG	Adoption of this technology is expected in 1999.
Geospatial Information Dissemination Services	Stream-based Model Interchange	RFP Released	OMG
Tagged Data Facility	Tagged Data Facility	RFP Released	OMG
Common Business Object Facility		RFI Issued	OMG
Shared Domain Components -- Electronic Commerce Domain	Negotiation Facility	RFP has been released	OMG

Applicable UTA Component	Applicable Specification	Status	Responsible Organization(s)	Comments
Geospatial Annotation Services
Geospatial Feature Analysis Services			OGC
Feature Generalization Services			OGC
Image Map Generation Services			OGC
Image Synthesis Services			OGC
Image Geometry Model Services			OGC
Geospatial Information Extraction Services			OGC
Geospatial Symbol Management Services			OGC
Image Understanding Services			OGC
Data Interchange Facility
Imagery Compression Facility
Information Storage and Retrieval Facility
Internationalization and Time Operations Facility
Mobile Agents Facility
Rendering Management Facility
Security Administration Facility
Shared Domain Components -- Electronic Commerce Domain	Asset and Content Management	RFI has been issued	OMG
Shared Domain Components -- Electronic Commerce Domain	Electronic Commerce Enabling Technologies and Services	RFI has been issued	OMG

6.4 Collaborative Computing

Collaborative computing is typified by the capability of one or more individuals and/or applications programs to share information in a timely manner. This section discusses these aspects of collaborative computing:

• Introductory description of the technology
• Geospatial requirements for this technology
• Standards that currently exist
• Available products that conform to the standards
• Status and future directions

6.4.1 Description

Examples of collaborative computing include the following:

• Simple file sharing (FTP, file sharing, etc.)
• Electronic mail (e-mail)
• Concurrent editing (e.g., shared whiteboard)
• Desktop audio teleconferencing
• Desktop video teleconferencing
• Data conferencing

Today, in a client/server environment, most collaborative computing takes place through the sharing of textual or formatted data that is exchanged via a file, database, or electronic mail server. Only occasionally does the information being shared take the form of image, graphic, audio, or video formatted data. This does not, for the most part, occur in real-time or anywhere near it, although some networks provide the capability for individuals to converse using text messages.

Elements of Internet and Java are now being incorporated into collaborative computing. The technology is reaching the point where it allows sharing information of all data types. The technology is emerging that will economically provide users on a network the ability to sit at their workstations and communicate via video and audio in real-time. In effect, a group of users will be able to create a virtual conference room, with the ability to share all forms of communication and information on demand.

Data conferencing is a relatively new form of collaborative computing. It is also called audiographics. Forms of data conferencing include shared whiteboard, document conferencing, and application sharing. Like other forms of teleconferencing, data conferencing is characterized by real-time communications, which differentiates it from "over-time conferencing" such as e-mail, newsgroups, and Lotus Notes.

6.4.2 MCG&I Requirements

Exploitation and Analysis is a key USIGS function that will make use of collaborative computing. Collaborative Exploitation and Analysis includes the sharing or the assisted performance of MCG&I exploitation/analysis tasks in two basic modes:

• Within peer group: Collaboration among a dynamically-defined analytical peer group, distributed organizationally and geographically
• Across functions: Collaboration in support of the collection/analysis tasking, data management, analysis refinement, information generation, and end-user feedback.

Collaborative exploitation and analysis includes workgroup-capable analysis, documentation applications, collaborative workspaces, intelligent agents to mediate, shared digital light tables, virtually-shared whiteboards, and collaborative planning/editing/production tools.

The vision for collaborative exploitation and analysis within the imagery & geospatial community includes these elements:

• A worldwide network of MCG&I data repositories that enable the sharing of MCG&I data and metadata across the community
• Worldwide, dynamically-formed workgroups of analysts, providers and consumers to address time-sensitive issues
• Cooperation of individual MCG&I analysts on joint exploitation tasks in real-time
• Access to MCG&I expertise
• Near real-time cooperation between analysts and end users for update
• The use of intelligent agents capable of browsing the MCG&I network and repositories for metadata information to support particular exploitation/analysis tasks ("pull")
• Automated, adaptive user profiling for data mining ("push") for the analysts
• The use of advanced visualization tools to achieve a "ubiquitous computing" environment.

6.4.3 Standardization

There are two main families of standards supporting collaborative computing: H.320 and T.120. H.320 is a family of video conferencing standards adopted by the ITU-T (the International Telecommunications Union Telecommunication Standardization Sector) that run over a variety of communications lines (T1, F-T1, ISDN BRI or PRI, Switched 56). The H.320 standards describe the frame structure and terminal procedures for multimedia communication multiplexed over one or more digital channels.

T.120 is a newer family of standardized communications and application protocols that provide support for real-time, multi-point data communications. It supports many kinds of multimedia data: images with real-time annotation, application sharing, and file transfer. These multi-point facilities are the building blocks for a new range of collaborative applications.

The DoD JTA, Version 2 [JTA98], has this discussion of emerging Video Teleconferencing (VTC) standards:

6.4.4 Products

Many products and tools exist for audio conferencing; shared whiteboard, video conferencing, and session management. For example, Compression Labs Inc., PictureTel, VTEL, British Telecom, Tandberg, Hitachi, Panasonic, and Mitsubishi all provide products that are H.320 standards-compliant. The T.120 standard is implemented in products from DataBeam, PictureTel, VTEL, British Telecom, Intel, Apple, Polycom, Vivo, Sony, Sun microsystems, GPT Video, Microsoft, and many more. The T.120 market leaders are PictureTel (Live PCS 50 and PCS 100) and Intel (ProShare Video System 200).

6.4.5 Status and Directions

A combination of audio-, video-, and data-conferencing is a feasible alternative today, in terms of cost, quality, and effectiveness. The technology does not yet provide an ideal solution in any of these areas, but it provides an adequate one. There are still a number of deficiencies to be noted:

• Scalability of virtual workplace tools to very large, distributed information environments remains an open question, both from time responsiveness and avoidance of overload
• More work is needed in integration of tools, to move from tool to tool and combine the analysis results. Collaborative tools need to be integrated with search and retrieval tools.
• Adaptation and integration of collaboration technologies for use by MCG&I analysts are unproven within the community
• Audio and video conferencing tools provide limited interoperability
  • Limited support for standards-based codecs
  • Limited support for the same codecs across audio and video tools
  • Announcements for future support for T.126 (Still Image Exchange) from many vendors, but not here yet
• Some audio conferencing tools only support point-to-point conferencing between two people; awaiting third party MCUs for multipoint capability

One problem with interoperability is that even if vendors follow the standards, there is no assurance that solutions from different vendors will interoperate out-of-the-box. Some work is being done to address this issue. Building on the ITU open video-conferencing recommendations of the H.320 and T.120 families, the Versit H.320 Implementors' Agreement has been developed which interprets and clarifies the ITU H.320 Recommendation. Many companies reviewed and contributed to this agreement. The development of this agreement is an important step in achieving the interoperability that will enable the collaborative working environment.

Future directions

The general trend in collaborative computing may summarized as follows:

• Conferencing will be built into general applications
  • A combination of application sharing and distributed document/object technologies will be used
  • Applications will be conferencing aware and provide new features for group collaboration
• Standards will become increasingly important for application internetworking
• There will be smooth integration of synchronous and asynchronous messaging
• The use of LAN, telecom, and Internet will be transparent.

6.5 Security

The combination of a distributed object environment along with the variety of envisioned USIGS customer/user types introduces significant architectural challenges for USIGS security. The issues are being identified and documented in a separate study [Usec97], and therefore will not be delineated here.