The cloud architecture shows modular services layers, auditors, brokers and cloud service management in the service oriented architecture. Mobile phones, laptops, PDA etc. There is a sense of location independence in that the consumer generally has no control or over the exact location of the provided resources but may be able to specify location at a higher level of abstraction ex.
Country, state or data center. Examples of resources include storage, processing, memory, network bandwidth, and virtual machines. To the consumer, the capabilities available for provisioning often appear to be unlimited and can be purchased in any quantity at any time. Resource usage can be monitored, controlled, and reported, providing transparency for both the provider and consumer of the utilized services.
Blower implemented web mapping in SaaS model on Google Application Engine and shows that web-mapping applications can be scaled well in the cloud computing environment. Mola et al. Mohapata et al. There is only a handful of literature reviews that exists in the domain of spatial cloud computing. Most of the literature identifies spatial cloud computing as the new research paradigm for GI science and emphasizes broader interdisciplinary research and collaboration.
Yang et al. The geographic perspective of cloud computing and spatial analysis of a cloud brings GI Science towards the unique frontier of computational science. The elastic nature of resources changes because of its invocation by different geographically dispersed users located in geographically dispersed computing environment brings the problem domain towards GI Science principles. The understanding of spatio-temporal characteristics will help to shape the cloud computing architecture and the services it provides.
The quality of services of LBS is inherently dependent on spatio-temporal locations of data centers, query locations, queried objects and end users locations. Schaffer et al. The authors also illustrated the importance of standardization and the role of OGC for interoperability of GI services in the cloud environment.
Cary et al. There has been a growing interest in the geospatial industries to enable the spatial cloud computing.
Recent technical white papers published by industry leaders such as Amazon and Intergraph identify the importance of cloud computing model for geospatial industry.
Implementation Models The deployment models of a cloud can be broadly divided into public, private and hybrid clouds. Public clouds are publicly available, mutli-tenant pools of computing resources that build on economies of scale paradigm. These clouds are more secure and private. GI services can be implemented in any of these models according to the organizational goal and availability of computing resources.
Exploring the GI services into various cloud models private, public, and hybrid also helps to unravel the potentiality of cloud computing for geospatial landscape. The GI services in the cloud in reference of licenses can be broadly divided into two main categories namely proprietary and open source models.
Open source software is usually developed as a public collaboration and often freely available. Open source refers to software system whose source code is available for use or modification by third-party developers; the term was adopted by a group of people on January at free software movement in California. The open source applications are available as cloud software, system software, desktop software, applications software, etc.
Cloud computing frameworks like Hadoop provide a platform for managing very large scale data often called big data. Other open source cloud framework suites include Eucalyptus, Nimbus, and Open Nebula.
Django-python has been used widely for PaaS, specifically for web application development in open source for GIS applications. Rodriguez- Martinez et al. Spatial Queries Conceptually, geographic data have been divided into vector and raster types. Both represent different ways to encode and generalize geographic objects and phenomena. The spatial queries use a lot of vector and raster data that are periodically generated through sensors, satellites, and GPS devices.
The large size of spatial repositories and complexities of geospatial models require upper time complexity in computation. The properties of spatial data make the spatial queries explicitly complicated. Pre- processing the data, analyzing the spatial data evolution, pattern identification and modeling are a critical component that makes spatial queries very complex.
Domain specific knowledge is inevitable to identify patterns of interest Han et al. Also, there has been substantial work in progress for spatial data mining query languages and geographic visualization. Furthermore, different techniques have been employed to improve the efficiency and representation of raster and vector queries. Spatial databases have been explicitly studied and different Spatial Access Methods SAM has been proposed in the past. Hierarchical data structures are predominantly used for representing spatial data structure.
R-tree proposed by A. Aguttman is spatial data structure that uses the concept of Minimum Bounding Rectangle MBR and has been widely used in different commercial and open source spatial database management systems.
Brinkhoff et al. Parallel Processing can be used to improve the performance of spatial queries that are non-linear. Papadopoulos et al. Further, Papadopoulos et al. Koperski et al. Carey et al. Zhang et al. Wang et al. Chiu et al. Karemi et al. However, there has been little progress in designing and implementing middleware for performing spatial analysis. The client server architecture during the initial development phase focused on implementation in the server-side.
However, the client side could also yield overall performance enhancements of these queries Vatsavai et al. Security and Interoperability LBS have a deep root beyond the technical scope of its implementation as a service. The implementations bring some of the social externalities as LBS by virtue depend upon the location data either for an operation or a legislative requirement. Hence, the applications of LBS that are used to monitor behavior of humans or resources and enhance participation could lead us to an era of surveillance.
Onsrud et al. The technical dimension of security and interoperability is again a key issue in distributed GI services like the cloud computing environment. This could be an exciting research opportunity that the GI science community should address in the future. The tools and framework needed to analyze these large spatial data sets are inexorably complex. Besides analyzing the architectural design issues, the second objective of this study was to design a prototype to enable Web Mapping Services WMS in a cloud using open source infrastructure.
However, the open source paradigm, which is a pragmatic methodology for free distribution of software codes and design, has been chosen to enable WMS in a cloud. It is trivial that each of the configuration files of all modules can be fully optimized and customized to our need for better performance of the system. Tools The prototype was implemented using various open source modules. Table 3. As the prototype is completely developed in the open source environment, most of these tools were downloaded from the internet and installed from a source codes.
The documentation that comes with most of the open source software was used to install and customize the modules. All feature-types were used to enable the web mapping application in the prototype. The boundary data was provided in polygon feature type that has six different attributes with only one record set.
The roads data were provided in a lines feature type with 50 different attributes and different record sets. Different sizes of the databases were used that varies with the number of attributes, record-sets and feature-types. All geometric feature-types were used in the prototype.
The shapefiles were converted to the physical SQL schema using the shp2sql tool. The cloud database was created in PostgreSQL server and all the schemas were imported in a cloud database. All the database schemas that were created for the prototype are listed in Appendix A.
Figure 3. Process used to manage the spatial data in the prototype. The shape file was downloaded from the city of Chicago web site. Shp2sql tool was used to create an intermediary SQL schemas, the schemas were finally imported to the postgreSQL database management system. The architecture of the system is illustrated in Figure 3. Design architecture of the prototype showing the key open source modules. The dotted line shows the interaction among the key components to publish a map in a cloud.
The web-mapping prototype was developed in the open source framework using Ubuntu as a system operating software. The python was used as a programming language for a development framework. PostGIS, the spatial component of the prototype was enabled on top of PostgreSQL to support the geographic data types and spatial functions.
Map Server, the image rendering engine acts as a middle tier web-mapping component to publish the map in a cloud through an Apache web server. Apache web server also served as a middle-ware to host web mapping application in the cloud.
All of these modules were downloaded from the internet with their source code and were compiled using gcc compiler to optimize with underlying hardware architecture and operating system. It ensured that the web mapping prototype was optimized for underlying hardware resources. For authorization purposes, the username and password should be verified to the PostgreSQL to query and publish the map. The user requests a map in the cloud through a browser provided to the user.
The requested map would be generated on real-time from the data stored in PostgreSQL server through the Apache server and is rendered by a map server. The different map files that are read by map server were created in advance. Those map files have granular information specified through a map class written in a structured semantic.
Map files constitute the detail technical specification of maps such as color, pixels size, extent, cartographic elements etc. Five different map files were created for the prototype. The lists of map files created and used in this prototype are listed in Appendix B. The python codes served as a backend programming environment and uses Application Programming Interface API among these modules in the prototype. Appendix C lists all the python codes written to publish a map in the browser in a cloud.
GUI can be accessed through network for processing the spatial queries to generate map files. The omnipresent of spatial data and ubiquitous computing led to tremendous increase in a usage of services like social networking, Location Based Services, and navigation systems and vice versa. Implementing Web Mapping Services WMS in a cloud infrastructure in an open source framework provides the challenges along with opportunities to leverage the value of service oriented architecture for GIS applications.
The prototype was developed using various loosely coupled modules. Each of the modules and components that were used are developed by different open source communities with their different design goals and objectives. However, by using all the modules in the prototype environment has illustrated the abstraction of encapsulated service of complex web mapping services.
The prototype can be analyzed as a Software as a Service and Software as a Platform paradigm in a cloud computing model. The design and architecture of the prototype was limited to specified map productions but the prototype is highly scalable to the larger production environment provided ample computing and communication bandwidth. Figure 4.
The prototype can be examined further on level of services it provides to the higher and lower layers of its SOA and also to the users of the prototype. The users can only published the specified map in this layer. The services it provided can be analogous to Software as a Service meaning that only specific maps can be published by the prototype.
Apache is an interface between Map server and a GUI. The users can use the map server to query the database and publish a map using a specific map file. Map files can be created by the users themselves if the users can logon to the system remotely using remote clients with administrative privileges. This is the lower level of the services provided by the proto-type where we can even specify the feature-types and data-types of the schema. Prototype in layered architecture.
The dotted line shows the different hierarchy of the different modules. The arrow line shows the interaction among key modules of the prototype. The designed prototype also illustrated how web mapping applications can be enabled in a cloud computing environment in Platform as a Service model and can take advantage of layered SOA architecture of the cloud.
Schaffer et. The prototype designed also used middle- tiers Apache web server and Map server illustrated on Figure 4. Yang et. Further, the prototype was fully developed using open source environment. Sieber et. The literature clearly illustrated that the cloud computing tiered architecture represents the best computing model for the GI science applications including web-mapping applications.
Further, the prototype is a Platform as Service models of a cloud reinforce the importance of modular components for scalability in a cloud environment. The files generated in prototype could be easily ported from the prototype to another compatible system and that is interoperable among multiple systems which is not possible in Google Application Engine environment.
The prototype can be further scrutinized to the n-tier architecture of a cloud computing environment proposed by NIST The access layer can be associated as GUI, which is analogous to access layer of the layered architecture of the cloud. The Apache and Map Server serves as a middle-tier layer in the architecture with map server being the key engine for web- mapping, which acts as an interface between the Apache web server and PostgreSQL server.
The map server and database server are a scalable component meaning that they can handle large sets of spatial data and render data into the web to generate the requested maps.
The SOA of the prototype can simply be exposed as each of the components are loosely coupled and provides the specific services to the other components and to users. The map server generates the map file. The apache server publishes it in the web. Furthermore, the prototype gives a notion of an abstraction where the underlying complexity is hidden to provide the web-mapping services. The map generated as the road map of the city of Chicago road networks in a line features-type.
Due to the large size of the records in roads database, there is a some delay to render the map in prototype system. However, the delay can be improved by increasing the physical hardware infrastructure of the prototype. The detail technical specifications of the generated maps in the prototype depend upon the map files read by the map server. The map files helps to customize maps generated by the prototype. Due to the small size of spatial database with polygon feature type, the map was published instantly.
Area, City of Chicago. The map constitute of two different layers. The CTA bus routes and boundary of Chicago. The single map file can generate multiple layers of map. Since there are more than eleven thousands CTA bus-stops, the maps points look like a line feature type. Although, the map rendering time increases proportionally with the number of rows and number of layers, the prototype published the multi-layer map in less than 2 seconds in a server.
The prototype will slow down if there is the bandwidth limitation when we are querying a large spatial database like roads. It clearly illustrates that web mapping applications needs more bandwidth to transfer a map in GUI of the client browser. The bandwidth can also be scaled as the Infrastructure as a Service in a network access layer in a cloud computing model but this is the limitation of the prototype.
All layers, City of Chicago. Nevertheless, the prototype system designed gives the notion of implementing web mapping applications in a cloud computing environment and supports the scalability of the system in terms of users, services and architecture.
The prototype is limited to web mapping applications in a cloud computing environment. The prototype development platforms are well tested in commercial operational environment. The prototype is limited to understand the architecture of web mapping applications in a cloud, thus, emphasized on the modularity of the open-source environment for scalability and optimization.
HPC in a cloud could be other options to analyze the large data sets in a cloud computing environment. The operational cost of the open source framework might be a financial setback instead of using metered proprietary services. The recent development of Kingston Automated Geoinformation Service KAGIS in illustrates how geo-processing in a cloud was successfully enabled in a national level in United Kingdom as a fully integrated service system.
On enabling spatial cloud computing, most of the literature reviews only calls for the broader collaboration among GI science communities that showed spatial cloud computing is still in infancy.
The GCI framework proposed by Yang et. However, there has been little progress beside a lot of awareness with limited collaboration and a few success stories on the research spectrum of spatial cloud computing. Moreover, the pertinent problem of interoperability and security in a cloud would always be pertinent in the GI applications too.
Nevertheless, the OGC framework has further pushed stakeholders for uniformity of spatial data and towards the cloud computing framework. Further, the transportation of large volume of spatial data sets from users to the providers could also be a key critical issue that has been largely neglected on most of the literature.
Beside technical limitation of cloud computing itself, the security and privacy in social dimension should be critically analyzed in the future as location data are highly sensitive. The prototype designed clearly illustrated that WMS services can be enabled on the cloud using open source framework in the cloud computing environment.
The prototype emphasized on neutral middle- tiers that are critical for interoperability in SOA. The cloud becomes a perfect computing framework to implement large scale web-mapping applications due to its SOA.
Further, it also supports the elasticity of Web-Mapping Application in a brink of Platform as a Service model in a cloud. Implementing the large scale GI services besides web mapping requires high level of management and collaboration in multiple tiers of a cloud computing framework.
Additionally, GI Science applications are inevitably interdisciplinary, so addressing the collaboration and work flow in GI Services and enabling spatial cloud computing is exciting research dimension.
Due to the requirement of the large volume of spatial data and their inherent association, cloud computing by virtue of its characteristics provides a promising framework for handling large scale GIS applications In addition, the geographic perspective of cloud computing is emphasized.
As the USDOL recognized geospatial industry as one of the fastest growing industries, the growth of geo-spatial data is inevitable. With the inception of cloud in geospatial landscape, further regulatory constraints will ease as cloud computing grows in acceptance.
With only a handful of cloud GI service providers in the market in geospatial landscape, developing GI science applications beside web mapping services in a cloud is still in its infancy and needs further scientific attention for the advancement of GI science.
GIS: Where are the frontiers now? Flag for inappropriate content. Download now. Related titles. Carousel Previous Carousel Next. Jump to Page. Search inside document. Netflix New and more efficient business models have emerged, making previous models increasingly obsolete.
How does it affect you? Esri Yudi Zulkarnaen. Sumit Gaware. Sharifa McLeod. Tyrese Hall. Jeffry Karundeng. Michael Valves. Arvin Edsel Sioson. Oke Adasen. Ankit Sharma. Dalibor Sopina Duc. Teguh Reinaldo. Diego H Rojas. Jessie James. More From John Michael Apacible. John Michael Apacible. Popular in Information Science. Van Rey Mercado. Talia Yvone. Tilahun Girma. Sree Gowri. Bashanta Dangal. SriRam Vadhul. Madya Imran Hafiz. Vamsi Karthik.
Surrogate Key Generator Implementation In datastage.
0コメント