Variable degrees of multimedia implementation and their impact on network elements

  • Authors:
  • Richard Bence;Jody Fraser;Laura Linser

  • Affiliations:
  • -;-;-

  • Venue:
  • SIGDOC '92 Proceedings of the 10th annual international conference on Systems documentation
  • Year:
  • 1992

Quantified Score

Hi-index 0.00

Visualization

Abstract

In developing an integrated document library for a NeXT LAN, we have experimented with multimedia applications and their implementation. In this paper we present results of this work including: issues of compression of different data types (e.g., voice, video, compound documents); demands placed on network bandwidth by multimedia documents in real-time, interactive use; processing loads generated by multimedia interactive use and impacts on perceived system performance; compound document and dynamically linked compound document file systems and their intra and internetwork transport; and real-time multimedia presentation durations determined by constraints of processing/storage requirements.Our experiments consisted of systematic measurements of multimedia specific packets on the 10mb ethernet bandwidth and ethernet to ethernet via routers and FDDI backbone. In addition, we collected data on processor utilization by specific tasks of the multimedia applications and correlated these with traffic measurements.We empirically studied the impact on network bandwidth of multimedia implementations to evaluate the need for intra-network compression and the trade-offs between transmission times and processing requirements for dual-end compression/decompression approaches.From this work we, developed guidelines and specific requirements for the NeXT platforms used in our testing with respect to memory, magnetic and optical media, display technologies and other factors such as dual end compression. We also developed a model which allows us to predict with some accuracy expected network loads, the processing demands placed on workstations and servers by multimedia implementations of different descriptions, and means of enhancing the performance of the system to accommodate these demands.These results provide a means of determining the hardware and software performance and configuration requirements of a specific type of multimedia implementation and, therefore, of the applicability or feasibility of multimedia technology for a particular application.As an example, library tutorial applications are high-level, focused on the look and feel of the library system, and make use of multimedia tools to convey these characteristics without placing large demands on the workstation processor server processor, or bandwidth. On-line help for the library system is more detailed, solution-oriented, keyed to components of the application suite and conventional in approach except for the addition of hypertext features and limited graphics. The trend towards inclusion of multimedia documents in growing numbers and increased complexity within the library system will affect it significantly without improvements in performance and management of resources.