《系统工程》课程教学资源（英文文献）The Design and Implementation of 3G Enterprise Logistics Management System

The Design and Implementation of 3G EnterpriseLogistics Management SystemAbstract:Todaywith the rapid development of computer and information technology,companiesinevitably have to make the transition from their old way in doing logistics management, which reliesmostly on humans, to a more computer-centric way, which relies mostly on logistical system softwareWith such trend,thereis anincreasingdemand forbuilding customized logistical system softwareacross many different industries. In this paper, we proposed a general framework for engineering suchsoftware system. Our framework is based on the MvC (Model-View Controller) software architectureIn our framework,the Model component represents the underlying data in the system, the Viewcomponent takes care of the display of the data; the Controller component manages the connectionbetween the Model and the View components. Our design guarantees the safety and integrity of thedata. The final system is portable, easy to maintain and operates efficiently. Towards the end of thepaper, we described how to further improve our logistic system to support 3G wireless users by usingdynamical web page generation technology.I.INTRODUCTIONIn today's market, the competition among enterprises become more and more intense.In order togaintheadvantageover other competitors, eachcompanyhas tokeepdeveloping newproducts topenetrate into emerging market area, as well as drive down the cost and increase the efficiency ofcurrent production to achieve higher profit. To achieve these goals, the companies need to accuratelygather information from customer market to internal production information, and process theseinformationquickly.Logisticmanagement system,which manages the whole supply/production chain ofthe companybecomes more and more important from the company's strategic perspective because the wholelogistics process covers lots of activities including purchasing raw material, making product, and storeproduct, take order and transfer/deliver products etc. At a high-level, the logistic activities inaccompany can be classified into five categories:(1) supply logistics; (2) production logistics; (3)saleslogisticsand(4)recyclelogistics and(5)disposelogistics.Withthedevelopmentof informationtechnology and increasing popularity of e-commerce, many companies have transitioned to rely onlogistical system softwareforlogisticmanagement.The ability to effectively and efficientlyprocessthose information becomes a key differentiating factor for companies in their industry. By improving

The Design and Implementation of 3G Enterprise Logistics Management System Abstract：Today with the rapid development of computer and information technology, companies inevitably have to make the transition from their old way in doing logistics management, which relies mostly on humans, to a more computer-centric way, which relies mostly on logistical system software. With such trend, there is an increasing demand for building customized logistical system software across many different industries. In this paper, we proposed a general framework for engineering such software system. Our framework is based on the MVC (Model-View Controller) software architecture. In our framework, the Model component represents the underlying data in the system; the View component takes care of the display of the data; the Controller component manages the connection between the Model and the View components. Our design guarantees the safety and integrity of the data. The final system is portable, easy to maintain and operates efficiently. Towards the end of the paper, we described how to further improve our logistic system to support 3G wireless users by using dynamical web page generation technology. I. INTRODUCTION In today’s market, the competition among enterprises become more and more intense. In order to gain the advantage over other competitors, each company has to keep developing new products to penetrate into emerging market area, as well as drive down the cost and increase the efficiency of current production to achieve higher profit. To achieve these goals, the companies need to accurately gather information from customer market to internal production information, and process these information quickly. Logistic management system, which manages the whole supply/production chain of the company, becomes more and more important from the company’s strategic perspective because the whole logistics process covers lots of activities including purchasing raw material, making product, and store product, take order and transfer/deliver products etc. At a high-level, the logistic activities in accompany can be classified into five categories:( 1) supply logistics; (2) production logistics; (3) sales logistics and (4) recycle logistics and (5) dispose logistics. With the development of information technology and increasing popularity of e-commerce, many companies have transitioned to rely on logistical system software for logistic management. The ability to effectively and efficiently process those information becomes a key differentiating factor for companies in their industry. By improving

the performance of its logistic management system, the company can reduce the in efficiency in eachstep of the logistic activity and achieve higher efficiency in its business.For these reasons and hopes, many companies started to build their own customized version oflogistics system in different industries. However, building a good logistic management system is achallenging task. Though most logistic management systems share some common pieces, differentcompanies in different industries have different requirement for different business activities. A gooddesign of logistic management system should not only address the current needs but also take intoconsideration for future maintainability and extension. Another important issues of logisticmanagement system is to seamlessly integrate with the current technology, (i.e. web2.0 and 3Gwirelessnetworks),In this paper, we present a frameworks for design and build logistic management systems. Ourframework is based on the MVC (Model-View-Controller) software architecture.The Modelcomponent represents the underlying data in the logistic system, the View component takes care of thedisplay of the data; the Controller component manages the connection between the Model and theView components. The separation of data and view architecture guarantees the safety and integrity ofthe data. The final system is portable, easy to maintain and operates efficiently. In order to leveragethe 3G mobile networking, we described how to further improve our logistic system to support 3Gwireless users by using dynamical web page generation technology. The following of the paper isorganizedasfollows.Section II describes the architecture of the logistics management system. Section III presented adetailed case study on how to design and implement the logistics management software. In section IVwe study the performance of our prototype system. Finally, section V concludes the paper anddiscussesthefutureworks.II.LOGISTICMANAGEMENTSOFTWAREARCHITECTURELogistic management software system can be classified into three parts according to itsfunctionality: warehouse management; order management and transportation management. Warehousemanagement keeps an inventory of all the items in the warehouse. It records the check in and checkoutactivitiesfor all itemsinthe warehouse.It classifiesthe items into different categories.Withthewarehouse information,the staff will be abletoknow whichitems are shortof need and whichitemsare surplus, and make the adjustment to optimize the storage for items. Order management includestaking order from the customer and dispatch the order to the internal business units for processing. Itmakes sure that the order are promptly and accurately processed. The major steps in order

the performance of its logistic management system, the company can reduce the in efficiency in each step of the logistic activity and achieve higher efficiency in its business. For these reasons and hopes, many companies started to build their own customized version of logistics system in different industries. However, building a good logistic management system is a challenging task. Though most logistic management systems share some common pieces, different companies in different industries have different requirement for different business activities. A good design of logistic management system should not only address the current needs but also take into consideration for future maintainability and extension. Another important issues of logistic management system is to seamlessly integrate with the current technology, (i.e. web2.0 and 3G wireless networks). In this paper, we present a frameworks for design and build logistic management systems. Our framework is based on the MVC (Model-View-Controller) software architecture. The Model component represents the underlying data in the logistic system; the View component takes care of the display of the data; the Controller component manages the connection between the Model and the View components. The separation of data and view architecture guarantees the safety and integrity of the data. The final system is portable, easy to maintain and operates efficiently. In order to leverage the 3G mobile networking, we described how to further improve our logistic system to support 3G wireless users by using dynamical web page generation technology. The following of the paper is organized as follows. Section II describes the architecture of the logistics management system. Section III presented a detailed case study on how to design and implement the logistics management software. In section IV, we study the performance of our prototype system. Finally, section V concludes the paper and discusses the future works. II. LOGISTIC MANAGEMENT SOFTWARE ARCHITECTURE Logistic management software system can be classified into three parts according to its functionality: warehouse management; order management and transportation management. Warehouse management keeps an inventory of all the items in the warehouse. It records the check in and checkout activities for all items in the warehouse. It classifies the items into different categories. With the warehouse information, the staff will be able to know which items are short of need and which items are surplus, and make the adjustment to optimize the storage for items. Order management includes taking order from the customer and dispatch the order to the internal business units for processing. It makes sure that the order are promptly and accurately processed. The major steps in order

management include taking the order, confirming the order and dispatching the order By querying thewarehouse information,the order management can reorganize the order and bundle similar ordertogether for dispatch. This will save the cost in order transportation. Transportation management is incharge of delivering the order to the customer. It will make the plan the for order transportation,choose the transportation route and pick the vehicle. A good transportation system will avoid waste ofspace in vehicle transportation and greatly improve the operating efficiency of the company.In order to optimallyachievethemanagement of alltheactivities,the logistics managementrequires a complex information system.The system should be able to process the information,managethe information and make decision based on the information. Processing the information includesgather the data, store the data, index the data and output the data. Managing the information includesorganizing the data and analyzing the data. The logistics management system will help the operator tomake the decision and plan for the futureControllerModelV个ViewFigure1Model-view-controllersoftwarearchitecture.Like any other software, the architecture is key to success of a logistics management system.Without a good software architecture, it is impossible to build an efficient system. In this paper, wepropose the adopt the MVC (model-view-controller) based architectureModetViewFigure 2MVC event handler process.Figure I shows the diagram for a model-view-controller based architecture. In this architecture,thesystem isdivided intothree components.Themodel componentmanagesthedata inthe systemThe view component manage the display of the data in the system. The control component manages

management include taking the order, confirming the order and dispatching the order. By querying the warehouse information, the order management can reorganize the order and bundle similar order together for dispatch. This will save the cost in order transportation. Transportation management is in charge of delivering the order to the customer. It will make the plan the for order transportation, choose the transportation route and pick the vehicle. A good transportation system will avoid waste of space in vehicle transportation and greatly improve the operating efficiency of the company. In order to optimally achieve the management of all the activities, the logistics management requires a complex information system. The system should be able to process the information, manage the information and make decision based on the information. Processing the information includes gather the data, store the data, index the data and output the data. Managing the information includes organizing the data and analyzing the data. The logistics management system will help the operator to make the decision and plan for the future. Figure 1 Model-view-controller software architecture. Like any other software, the architecture is key to success of a logistics management system. Without a good software architecture, it is impossible to build an efficient system. In this paper, we propose the adopt the MVC (model-view-controller) based architecture. Figure 2 MVC event handler process. Figure 1 shows the diagram for a model-view-controller based architecture. In this architecture, the system is divided into three components. The model component manages the data in the system. The view component manage the display of the data in the system. The control component manages

the logistics in the data. In this design, the complex system is decoupled into three subcomponent andeach subcomponent can be developed, tested and maintained independentlyFigure 2 gives the event handler process in the MVC model. First the user gives an event to thecontroller component. The controller component analyzes the event and translates the event into aservice. The controller component then sends the service request to the model component. The modelcomponent receives the request and performs the service to update the data. After the model updatethe data, the view component update the display.The goal of model-view-controller design to separate the data from the view. When the data andtheview are separated, there areless coupling betweenthe codes.Andthe codebecomeless complexand easier to maintain. It also increase the flexibility of the whole system architecture.In thisarchitecture, it is very easy to swap in new display front-end into the system without completelyredesign the whole system. A good example usage of this is to support different devices to access thesystem.By swapping different view components, the system can support both desktop computer withEthernet cable, notebook with wireless networks and smart phone in3G networks. One of a goodchoice of the frontend view component is to use HTML/XHTML to display the data. With thisarchitecture, the user doesn't need to install any software. All the operation are done through thewebpage. This not only make the system easier to the user but also is independent to the Os for thefrontend device. Another major benefit is that system update becomes very easy. We only need toupdate the sever end and requireno update on the user sideIII.CASESTUDYIn this section, we use the logistic system design for concrete company as case to study the MVCsoftware architecture. The main focus for this system is to process online orders from the customerThe customers browse the website and look for the concrete information. Once they find the type ofconcretetheywant,theywillgothroughtheonlineorder system inthelogistic webpagetoplacetheorder. After the order has been placed, the frontend sever receives the order and confirm the orderWith the customer. At the same time, the frontend will start processing the order. When the fund ispaid, the transformation department will start the whole process of logistics, such as dispatching theorder, packaging the item and arranging delivery the order. During this period, the system will timelyupdate the order status information and allow the customer to query the status of the order. The wholeprocessendswhen thecustomerreceivedtheorder

the logistics in the data. In this design, the complex system is decoupled into three subcomponent and each subcomponent can be developed, tested and maintained independently Figure 2 gives the event handler process in the MVC model. First the user gives an event to the controller component. The controller component analyzes the event and translates the event into a service. The controller component then sends the service request to the model component. The model component receives the request and performs the service to update the data. After the model update the data, the view component update the display. The goal of model-view-controller design to separate the data from the view. When the data and the view are separated, there are less coupling between the codes. And the code become less complex and easier to maintain. It also increase the flexibility of the whole system architecture. In this architecture, it is very easy to swap in new display front-end into the system without completely redesign the whole system. A good example usage of this is to support different devices to access the system. By swapping different view components, the system can support both desktop computer with Ethernet cable, notebook with wireless networks and smart phone in3G networks. One of a good choice of the frontend view component is to use HTML/XHTML to display the data. With this architecture, the user doesn’t need to install any software. All the operation are done through the webpage. This not only make the system easier to the user but also is independent to the OS for the frontend device. Another major benefit is that system update becomes very easy. We only need to update the sever end and require no update on the user side. III. CASE STUDY In this section, we use the logistic system design for concrete company as case to study the MVC software architecture. The main focus for this system is to process online orders from the customer. The customers browse the website and look for the concrete information. Once they find the type of concrete they want, they will go through the online order system in the logistic webpage to place the order. After the order has been placed, the frontend sever receives the order and confirm the order With the customer. At the same time, the frontend will start processing the order. When the fund is paid, the transformation department will start the whole process of logistics, such as dispatching the order, packaging the item and arranging delivery the order. During this period, the system will timely update the order status information and allow the customer to query the status of the order. The whole process ends when the customer received the order

CustomerwareHousesomgDINTWareHouse_has_TranspcrtationWersHuneDNSneVAROHA4S)DINTWHoeWaHoDINTSDLCICE VARCHAR(S)TrargportalenTranspaetationdDnI VARQ1AR(45)Addea APOHARt45dmOrdeTransportatonPordenprWarehuusefatirageCeottanlmPredtIDVIODenSiVARCHAR(45)WsrhousehorocelDDNCAmartDOLBLEsberuJtemYARCHAR(4S)DaeDATE9CostDOUDLEORemainingCapaxityIVTS:ab: WAFCHAP(45)Vahide,yehickDMetimakaoadteJMRetunOrder_BetumordgDINTPAredutPrautDNTProducthaswaielousProdudt_ProouctIDINTWe0mDeliverOrdervehideDeliverOrderIDDNTFXehideDINdnoRetumorderaVARCHARESOTO MARCHAR(45)TRetunOrderIDMTCSrVARCHWR(45)DAmixrtDoLBLECHgnaOrderIDARCHAR(A5)Capoty INTSTimeDATEO.From XARCHAR(40)rdicnUROH/45)urtDoUBLEProdkxSupplieProdutID UNTSulerDNTManeKARCHAR(45NOMEYARCHA(4S)OPICeDOLELhare VARQHA(45)VARCIOP(45)AJess VARCTAR(45)Email UARCHAR(45)aroHouse_VareHousatDINTFigure 3 ER model for data design.Figure3shows the datamodel designforthe system.Weusetheentity-relationrepresentationtodescribe the structure of the data. As seen in Figure 3, there are 12tables: customer information table,concrete information table, order table,vehicle management table,deliver order table, warehousestatus table,warehouse storagetable,transpiration table,product table,return order table, suppliertable.IV.EXPERIMENTALRESULTSIn this section, we present the experimental result for our prototype system.Figure Error!Reference source not found. Shows the platform configuration for the prototype system, whichinclude a back-end database server, a web server and a user front-end.UserServerDatabaseFigure 4 Configuration of prototype systemThereare twokinds of experiments for this softwaresystem.Thefirst kind of isfunctional

Figure 3 ER model for data design. Figure 3 shows the data model design for the system. We use the entity-relation representation to describe the structure of the data. As seen in Figure 3, there are 12tables: customer information table, concrete information table, order table, vehicle management table, deliver order table, warehouse status table, warehouse storage table, transpiration table, product table, return order table, supplier table. IV. EXPERIMENTAL RESULTS In this section, we present the experimental result for our prototype system. Figure Error! Reference source not found. Shows the platform configuration for the prototype system, which include a back-end database server, a web server and a user front-end. Figure 4 Configuration of prototype system. There are two kinds of experiments for this software system. The first kind of is functional

experiments Functional experiment can be done manually and automatically. Most the data integrityand completenessexperimentcanbedoneautomatically.By comparingthegoldendata withthetruesystem output, we will be able to verify the correctness of the system. Most of the user interface anduser interaction has be done manually.The user interact with the system through the interface.Themain purpose of manual test is for the appearance of the control such as shakeup, color, size etc.The second kind of experiments is to study the performance of the system. There are twoimportantPerformance characteristics for the system: (1) throughput and (2) latency. Throughput describethe maximal data the system can process in a given time. While the latency describe how long it takesfrom the initiation of the request to the final received of the data. Both measurement are important.Thehigherthethroughput, themorecustomerthesystem can serve.The shorterthelatency,thelesslikethecustomergoesawayElansedTime InTest (seeslFigure7Plotofresponsetime.In this experiment, we wrote a simple http data connection program to randomly connect to theonline system and simulate the order submission.The simple http connection program also records thesuccessor or failure of the order submission with the response time. Figure 5shows a snapshot the rawresponse time record in the simple data connection program. In this snapshot, there are two phase ofconnection. The first 20 seconds, there are large variation in the response time. The networkingcondition is not very stable.After 2o seconds, the response time stays very stable even when theconnectionfrequencyincreases.Thenetworkingconditionsimprove

experiments Functional experiment can be done manually and automatically. Most the data integrity and completeness experiment can be done automatically. By comparing the golden data with the true system output, we will be able to verify the correctness of the system. Most of the user interface and user interaction has be done manually. The user interact with the system through the interface. The main purpose of manual test is for the appearance of the control such as shakeup, color, size etc. The second kind of experiments is to study the performance of the system. There are two important Performance characteristics for the system: (1) throughput and (2) latency. Throughput describe the maximal data the system can process in a given time. While the latency describe how long it takes from the initiation of the request to the final received of the data. Both measurement are important. The higher the throughput, the more customer the system can serve. The shorter the latency, the less like the customer goes away. Figure 7 Plot of response time. In this experiment, we wrote a simple http data connection program to randomly connect to the online system and simulate the order submission. The simple http connection program also records the successor or failure of the order submission with the response time. Figure 5shows a snapshot the raw response time record in the simple data connection program. In this snapshot, there are two phase of connection. The first 20 seconds, there are large variation in the response time. The networking condition is not very stable. After 20 seconds, the response time stays very stable even when the connection frequency increases. The networking conditions improve

.-.nFigure5Plotofstatisticsofresponsetime.Figure 6 shows the statistics of the response time. There, yellow and green lines represented 90%,80% and averageresponse time, respectively.Overall the response of the system is pretty fast (lessthan 0.8 seconds). Figure 7shows the throughput for the system. When the connection frequencyincrease to the most around 25 seconds, the system achieve a maximum throughput of50order/seconds.sedTenTet/serFigure 6 Plot of throughputV.CONCLUSIONIn this paper, we presented a software framework for designing 3G enterprise logisticsmanagement system.Our framework is based on the model-view-controller (MVC)architecture.Inour framework, the Model component represents the underlying data in the system; the Viewcomponent takes care of the display of the data; the Controller component manages the connectionbetween the Model and theView components.By decoupling the data from the view, the systemdesign become less complex and more maintainable. Our design also guarantees the safety andintegrity of the data.This also support multiple frontend devices from desktop computer to 3Gsmartphone by using dynamical web page generation technology.With this framework, we developed a prototype online order logistic management system for aconcrete company. The outcome system is portable, easy to maintain. Simulation results havedemonstrated fast response time and high throughput of the prototype system. In future, we are going

Figure 5 Plot of statistics of response time. Figure 6 shows the statistics of the response time. There, yellow and green lines represented 90%, 80% and average response time, respectively. Overall the response of the system is pretty fast (less than 0.8 seconds). Figure 7shows the throughput for the system. When the connection frequency increase to the most around 25 seconds, the system achieve a maximum throughput of 50order/seconds. Figure 6 Plot of throughput V. CONCLUSION In this paper, we presented a software framework for designing 3G enterprise logistics management system. Our framework is based on the model-view-controller (MVC) architecture. In our framework, the Model component represents the underlying data in the system; the View component takes care of the display of the data; the Controller component manages the connection between the Model and the View components. By decoupling the data from the view, the system design become less complex and more maintainable. Our design also guarantees the safety and integrity of the data. This also support multiple frontend devices from desktop computer to 3G smartphone by using dynamical web page generation technology. With this framework, we developed a prototype online order logistic management system for a concrete company. The outcome system is portable, easy to maintain. Simulation results have demonstrated fast response time and high throughput of the prototype system. In future, we are going

to study how to extend our system for multipledatabases and multiple severdesigns.An even moreinterestingstudyisthat boththe databasesand web servers arelocatedindifferentregions oftheworld

to study how to extend our system for multiple databases and multiple sever designs. An even more interesting study is that both the databases and web servers are located in different regions of the world