{"title":"An instance of the application download pattern: the SPAIDS software loader/verifier domain analysis and implementation","authors":"J. D. Riley, Sangiv Dungari, W. Pritchett","doi":"10.1145/269629.269656","DOIUrl":null,"url":null,"abstract":"The requirement to obtain application sofhuarefiom an external source and correctly load on a potentially embedded target is a COIWIIOI~ problem in many domains. For example, a growing solution in an internet environment is the use of a Java \" 'applet \" loaded to a target Java Virtual Machine (NM) implementation. Another applicable domain is embedded systems particularly Avionic and Vehicular Control software. Similar problems are COIIUIIOII to all these domains. These include accessing an appropriate communications medium, authenticating that the application software is virus and tamper free, and verifjng the load has occurred without error. We identii this-as the Application Download Pattern. In the Vehicular Electronics (Vetronics) field this problem is referred to as the So&are Loader/Verifier problem. We describe the SPAIDS Sofiare LoaderNerifier (Slav) project, place it in the context of the general application download pattern, and then relate lessons learned in the SLN analysis and implementation to the general problem. The domain analysis is expressed using the Unifed Modeling Language Version 1.0 (UML). The implementation directly expresses the UML model making substantial use of the Object Oriented features ofAda 95. 1. Background-The SPAIDS SUV Project When software was first introduced to Army Ground Combat Vehicles, an accessory, the Software LoaderNerifier (SIN) was introduced to load the battlefield system software into the vehicle. AS the Ground Combat Vehicle (GCV) domain has grown with various types of vehicles manufactured to support specialized missions, multiple versions of the SLN have emerged. Each type of Army GCV has its own SIJV associated with it. This has resulted in specialized training for each type of SLN, multiple baselines of software, high maintenance costs and a large number of total units to maintain. The objective of the SPAIDS SIJV project is to perform domain analysis of various platforms, to generate domain specific software and hardware reqnirements, and to design and develop a prototype SUV with reusable components to support the transition from platform-specific, SuVs to a platform-independent, generic (e.g. common software and hardware components) form. OfdlOrp~Oofthis~~O~f~ personal or cl;lssroml us0 is graatd without fee provided that c@= afe not ,,,& or di&b~ted for prolit or commercial advantage, the mPYr%@ neti=, Ihe tile of&e publication and its date appear, and nOti= b &'a ulat copying is by permission of ACM, lnc To CqPY othe> t republish, to posl on sewxs, or to redistribute to lu% re ¶ur@ Puor spific permission aedlor a f'ae. The …","PeriodicalId":429060,"journal":{"name":"TRI-Ada","volume":"14 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1997-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"TRI-Ada","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/269629.269656","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The requirement to obtain application sofhuarefiom an external source and correctly load on a potentially embedded target is a COIWIIOI~ problem in many domains. For example, a growing solution in an internet environment is the use of a Java " 'applet " loaded to a target Java Virtual Machine (NM) implementation. Another applicable domain is embedded systems particularly Avionic and Vehicular Control software. Similar problems are COIIUIIOII to all these domains. These include accessing an appropriate communications medium, authenticating that the application software is virus and tamper free, and verifjng the load has occurred without error. We identii this-as the Application Download Pattern. In the Vehicular Electronics (Vetronics) field this problem is referred to as the So&are Loader/Verifier problem. We describe the SPAIDS Sofiare LoaderNerifier (Slav) project, place it in the context of the general application download pattern, and then relate lessons learned in the SLN analysis and implementation to the general problem. The domain analysis is expressed using the Unifed Modeling Language Version 1.0 (UML). The implementation directly expresses the UML model making substantial use of the Object Oriented features ofAda 95. 1. Background-The SPAIDS SUV Project When software was first introduced to Army Ground Combat Vehicles, an accessory, the Software LoaderNerifier (SIN) was introduced to load the battlefield system software into the vehicle. AS the Ground Combat Vehicle (GCV) domain has grown with various types of vehicles manufactured to support specialized missions, multiple versions of the SLN have emerged. Each type of Army GCV has its own SIJV associated with it. This has resulted in specialized training for each type of SLN, multiple baselines of software, high maintenance costs and a large number of total units to maintain. The objective of the SPAIDS SIJV project is to perform domain analysis of various platforms, to generate domain specific software and hardware reqnirements, and to design and develop a prototype SUV with reusable components to support the transition from platform-specific, SuVs to a platform-independent, generic (e.g. common software and hardware components) form. OfdlOrp~Oofthis~~O~f~ personal or cl;lssroml us0 is graatd without fee provided that c@= afe not ,,,& or di&b~ted for prolit or commercial advantage, the mPYr%@ neti=, Ihe tile of&e publication and its date appear, and nOti= b &'a ulat copying is by permission of ACM, lnc To CqPY othe> t republish, to posl on sewxs, or to redistribute to lu% re ¶ur@ Puor spific permission aedlor a f'ae. The …