4.1 Agent-based systems
4.3 Provide a literature review on Agent-based systems
In 4th Generation (4G) Networks, deployment of multimedia applications have been effective in allowing users with ubiquitous computing enabled smart phones to transcend physical and social barriers to exchange information and connect with others anywhere across the globe. The 4G revolution has crucial antecedents of innovative service provisioning frameworks which take into account user buying behavior, device sophistication and available content providers. These dynamics, and others that shall be discoursed in this research paper, should be considered by network operators as they develop billing systems to preclude laborious computation of proceeds for entities involved in the provisioning practice. This paper proposes a system for automated business-driven service provisioning. Evidence of concept execution is set up in 4G mobile network. A smart system, capable of utilizing swaths of information, including operator’s corporate tactics, accessibility of new programs, content modules, user location, device specifications and the amount of credit on user account, is operationalized to help operators to facilitate on-the-spot generation of list of potential users of new services.
Keywords: 4G Networks, ubiquitous computing,
The past decade has witnessed an upsurge in technological mediation in business activities. This trend has been attributed, in part, to fast-paced growth and adoption of information communication technology (ICT) offerings. As a result, inherent benefits of ICT growth such as superior computing capabilities, vastly improved world-wide web, global internet access and connectivity, all noticeable antecedents of virtual reality, and business integration have gained prominence (Monostori, Váncza & Kumara, 2006). Similarly, rapid expansion and transformation of markets typified by high number of entrants into emerging sectors such as virtual enterprises, integration of processes, customer-driven logistic networks, and e-commerce has headlined demand for Artificial Intelligence (Vasirani & Ossowski, 2012; Luger, 2008). Whereas there are ongoing deliberations among artificial intelligence policymakers and researchers on defined path of technological dynamics of technologies, there is concurrence that accessibility of data and the capacity to proliferate, and derive meaning from it is idealized as indispensable asset for firms that seek to confront mounting customer needs in world markets. As a result, Agent-based calculations present new standard of computing expertise that fundamentally refines and enhances technological capacity to establish the aforestated patterns (Monostori, Váncza & Kumara, 2006). In recent history, Agent-based conceptualizations and practices have been discussed in myriad scientific and engineering literature. In practice, agents are expected to solve problems arising from “autonomy and complexity” and are as such expected to respond to vacillations and interruptions by demonstrating intelligence. In their computing roles, agents are perceived as social actors with the capacity to exhibit an ability to facilitate self-recovery and response to real-time ICT trepidations. Thus, Agents are a vital cog in the global ICT revolution, as it infers an expansive business stage, geographically and technologically. These and other conceptions have been adopted into industrial enterprises since they accelerate attainment of significant characteristics, namely, sovereignty, sensitivity, expendability, distributedness, and sincerity. Agents may possibly be deployed in situations in which there are undefined and inadequate data and facts. Therefore, numerous responsibilities associated with manufacturing, such as engineering design and supply chain management could be assigned to agents who should be authorized to interconnect and collaborate with similar entities.
This step focuses on Application of Agent Based Systems to Service Provisioning Problems in 4G Mobile Networks. Specifically, this section recommends a solution for a practical situation involving the creation of a list of potential users for the new service. The proposed methodology is anchored on the Business Manager Agent (BMA), an interface linked to the network provider’s Business Support System. Petric et al. (2007) offers a cursory description of the proposed solution by arguing that satisfaction of necessary rudiments of business from the user‘s end activates the Java Expert System Shell (JESS) rule engine codified as “http://herzberg.ca.sandia.gov/jess” which then notifies the BMA to induce the automated service provisioning process. Automated service notification is activated at this time, just about the same time a device-specific type of the software component is activated. At this stage, the service generates a report to the BMA (Petric et al., 2007; Dumic, 2007). To enhance data security, download of the software notification from the download server is deferred until a strategy is in place for its execution. Subsequently, the Software Retriever Agent (SRA) retrieves the software module as the Remote Management Shell (RMS) agents position it at the Broadcast/Multicast Service Center (BM-SC) nodes to facilitate retrieval in hand-held devices and subsequent execution on their mobile phones (Kusek, et al. 2003). The methodology also employs the Charging Manager Agent (CMA) which functions as an interface for the Billing Server. Upon confirmation of request from the BMA, the CMA counterchecks the balance on user’s account. Charging Manager Agent is guided by checks and controls specified in the BMA. The quantity of messages received at the CMA should reflect the magnitude of strategies to be implemented. Considering the fact that messages from the BMA are specific to target users, it is imperative for the CMA to utilize real mobile numbers, preferably, the International Mobile Subscriber Identity (IMSI) numbers. The latter are expected to conform to the requested type (Petric et al., 2007). Subsequently, the CMA connects to the Database Agent (DBA) which has privileged access to the database containing user profiles. The DBA inspects the database and returns the IMSI numbers for particular users assigned a given user type to the CMA as shown in Figure 1 below.
Figure 1. Finding the IMSI numbers (Petric et al., 2007)
The CMA authenticates the received user accounts to determine eligibility of subscribers to the service. Eligibility is usually determined on the basis of the amount of money in in a given user’s accounts as shown in Figure 2. Incidentally, the BMA is quite versatile and can be dedicated to optimizing sequence of implementation of business strategies
Figure 2. Checking the users’ accounts (Petric et al., 2007)
In this way, optimization is attained by through the genetic algorithm approach (Petric et al., 2007). Upon completion of sequencing of strategies, the BMA returns an outcome to the Provisioning Manager Agent (PMA). The PMA essentially coordinates the functions of SRAs, and are additionally applicable to real utilization procedure implemented through the RMS system. Petric et al. (2007) identifies the parameter as specifically containing “nature of content, location, time, QoS, user type” are necessary for location of target users. The identified users are delineated as those who have genuine interest in specific mobile and are within locations demarcated in the business strategy. Upon authorization by the BMA through an optimal execution order, the PMA signals the DBA to demarcate potential users and subsequently forwards the list of users concerned in accordance with their unique IMSI numbers as shown in Figure 2. It has also been proven that the CMA is essentially scalable and can be used to amalgamate both users’ lists for particular approach. Upon receipt of the ultimate list of targeted users, the PMA executes the service provisioning function on behalf of selected users. Principally, the SRA obtains a catalogue of software programs that are yet to be procured. Once procurement is certified as complete, the RMS agents can initiate “the migration and installation of that software on the remote BM-SC node.” Upon completion of installation the users are authorized to begin using the new service on their terminals by means of the Access Network.
This step explores research findings through theoretical analysis of the research topic based on the methods above. A number of exploratory studies favorably associate Business-driven Service Provisioning Process in 4G mobile networks with potential alteration of current configurations in the telecom industry by facilitating the direct connection of users to a single network provider (Petric et al., 2007). Since current stakeholders in 4G service network include, “network operators, content providers, independent software providers and users”, past research suggest the crowded field of actors, will precipitate emergence of new problems. Similarly, Kusek et al. (2003) positively correlates high number of stakeholders with billing challenges. Particularly, is projected that network operators and content developers will struggle to craft a suitable tradeoff between high unit costs and high sales volumes. Undoubtedly, the stakeholders will encounter stiff challenges in the determination of billing rates at which they need to charge mobile users for services in 4G mobile networks. Additionally, efficient billing of network services in “telecommunication systems architecture” will inevitably necessitate various kinds of nodes and mobile equipment. This underscores the necessity of determining client characteristics and behavioral properties, account balance on their devices and the user’s account details during the service provision process (Petric et al, 2007).
Similarly, strategies to increase revenue will almost certainly require augmentation of sales. The same effect may be achieved by instigating cost-reduction measures such as lowering operational costs. It is also logical for network operators to go beyond cost-cutting to employ innovative strategies to “satisfy new user demands”. This will likely compel the network operator to rationalize operations in accordance to the proposed business strategies. It has also been established that software agent strategic choices are mediated by the operators’ business strategies that mirror dynamics and permutations in the telecommunications sector (Petric et al., 2007). Moreover, several studies, notably by Petric et al (2007) and Kusek (2003) associate Agent strategies with positive outcomes such as sustainable competitive advantage. In this respect, competitiveness is attained by leveraging on advantages derived from having intricate knowledge of the consumer behavior and Business to Business (B2B) relationships in the external business environment. Standard practice by many operators reveals a general proclivity to for the “rule engine”, a framework by which defined business approaches are encased. Indeed, the implication of this strategic trajectory is for business strategies should to be encapsulated in firm wide rules that guide provision of new mobile services. Besides, the business rules that define official statements on business information and procedures. As typical multi-agent systems employ “action enabler rules” they tend to induce an action upon validating certain prescribed conditions as true (Petric et al., 2007). The enterprise rules encompass necessary parameters for describing new mobile services as may be delineated by nature of the service content, type of mobile device, service execution sites, category of the target users, and the software modules for the implementation of services on mobile devices, Quality of Service (QoS), and duration of execution of the chosen strategy (Petric et al., 2007). Additionally, other parameters are contained in a different location known as The Charging Server. This location stores relevant data concerning the client accounts and applicable tariffs for each of the services available to the customer. As a result, Agent-based systems are effective for delivering Content and Event based billing to the benefit of the operator and the content providers through the Service Charging Application Protocol (SCAP). Thus, flexible billing practices of multimedia services are executed in conformity to determined service charges. The “Provisioning portal” avails a user-friendly interface that displays user predilections, contextual data, and device specifications. On the other hand, the User Profile Directory contains user profiles suitable for ascertaining whether software upgrades are permitted. It also encloses a catalogue of International Mobile Equipment Identity (IMEI) numbers that distinctively categorize the type of mobile device used b specific users (Petric et al., 2007). The Terminal Profile Directory provides information on characteristics of various mobile devices based on their vendors. The Home Location Register (HLR) facilitates mobility management by availing information regarding the user’s location.
This section summarizes related work, compares methods and results from reviewed literature. Studies by Monostori, Váncza and Kumara (2006) have presented the Agent -Based standard of computing and deliberated the characteristic features of software agents as well as the procedural actions of multi-agent systems. From methodical survey of selected Agent Applications, the authors perceives various agent technologies as fundamentally indispensable in all facets of manufacturing since they facilitate realization of recommended Agent properties as “autonomy, responsiveness, modularity and openness” (Monostori, Váncza & Kumara, 2006). The study also scrutinizes multi-agent systems deployed in a decentralized manner and how they are assimilated to distributed and incomplete data and knowledge sources. On uncertainties and subsequent conflicts, the focus shifts to communication, collaboration and cooperation, as a strategy for circumventing such challenges. While the agent-based approach permits open-ended plans and execution of intricate systems, the study acknowledges a number of bottlenecks which typically manifest as challenges of scalability, safety and traditional software quality. The main impediments to global adoption of agent technologies in manufacturing industry have also been summarized. These include vacillations in international operations, incessant fights involving agents and self-seeking entities, and poor communication. In recent times, the industrial acceptance of Multi-Agent Systems (MAS) in engineering has been comparatively low, a situation blamed in part, to hitches arising from failed attempts to integrate with existent legacy systems. Integration of MAS and legacy systems persists even when incremental approach is preferred. However, advances in Agent technologies are regarded as exceptionally vibrant, pioneering and “ramifying” (Monostori, Váncza & Kumara, 2006). While still on the subject, Agents tend to take upon themselves a burning obligation to facilitate convergence with modern industrial software technologies. Henceforth, it is anticipated that agent-based computing will evade inherent challenges of Agent based systems. Accordingly, while drawing from qualitative research, Monostori, Váncza and Kumara (2006) identify the overarching goal of industrial manufacturing as the “general conversion of all capital to the satisfaction of human needs. Consequently, Monostori, Váncza and Kumara (2006) emphasize four critical directions: adaptive, digital, knowledge-based, and networked manufacturing as new paradigms for future manufacturing. The “four directions” are nevertheless comprehensively aggregated under the concept of cooperative manufacturing (Monostori, Váncza & Kumara, 2006). Still, all four directions remain indispensable to the realization of numerous agent technologies. It is projected that development of multi agent-based systems will likely be in tandem with manufacturing, though the former is projected to create more challenges for the latter.
Cross-sequential studies by Fisher (1994) employs Concurrent MetateM, a multi-agent programming language to derive meaning out of agent based systems through application of executable temporal logics, assimilated into broadcast message-passing and grouping mechanisms. Subsequently, their applications to diverse engineering contexts have been conversed and the approach vindicated in both “logical and practical terms” (Fisher, 1994). In addition to the concept of autonomous objects, Concurrent MetateM also offers a greater structuring mechanism by its ‘groups’ extension. This feature limits the degree of an object’s interaction, and permits additional “mechanisms” for the refinement of Agent-Based approaches for firms. In this way, it presents the Agent with the opportunities for realization of collaboration, rivalry and communication. In addition, Fisher (1994) has proved the potency of logic-based approach as a theoretical model and a practical modeling tool. Some scholars have revered the Concurrent MetateM language for its natural efficacy in capturing an extensive collection of agent behaviors. It is hoped that irrespective of the efficiency of the combined approach, the concepts of implementable agent characteristics and group-based agent interactions may be applied discretely in Future Systems. Finally, Fisher (1994) demonstrates the potential benefits of closing gaps between theory and practice. Also, the text has exposed the possibility of preserving relevant elements of agent description while opening new frontiers for future innovation.
Similar studies by Balakrishnan (2019) have investigated Agent Based Systems from the perspective of Intelligent Agent. The Intelligent Agent perspective is anchored on the belief that Agents “which surrounds the same purpose”, will aid the discovery of appropriate content in an effectual and efficient manner (Balakrishnan, 2019). In brief, the Intelligent Agent- based systems facilitate resolution of complex practical issues through intellectual cooperation to form a sustainable pathway for making stumbling blocks and subsequently group-up the outcome (Liu & Barber, 1998). All things considered, the IA agent-based approach has proved effective in aiding the development of lots of agent generation algorithms.
Similar works by Stewart (2009) on Agent-Based Systems have fundamentally focused on their applications with respect to possible effects of imposing numeric principles on a number of “states and actions” on “tractability” of basic success and maintenance of agent design problems. Essentially, it has been established that agent design problems remain intractable irrespective of the severity of numeric conditions. Accordingly, the aforestated conditions preclude tractability of agent design problems to the extent that a contrary outcome is only possible through “restrictions” on other sorts (Stewart, 2009).
In the same way, analyses by Winikoff and Cranefield (2014) found that the probability space of possible behaviors for Belief-Desire-Intention (BDI) agents is, in fact, large, when considered in both absolute and in relative terms. The preceding declaration is valid when paralleled with technical programs of the same size. In keeping with theoretical projections, it is anticipated that the number of possible behaviours will increase with corresponding increase in the tree’s depth and breadth (Winikoff & Cranefield, 2014). Conversely, to a certain degree, it is surprising that addition of failure handling causes a markedly substantial change to the “number of behaviours”. For example, for immutable goalplan tree 3 units deep and 2 units wide, introducing failure handling escalated the quantity of “successful” behaviours from 128 to 6,332,669,231,104 (Winikoff & Cranefield, 2014). Another advantage of BDI-mediated agent based system permit the delimiting of extremely flexible and robust benefits. These facts have been put forth in analyses by Winikoff and Cranefield (2014) who provide numerical justifications for long-standing persuasion that BDI agents permit for the description of highly flexible and robust agents. Flexibility in this regard is delineated as the sum total of probable behavioral attributes of an agent, which has been proven to be sufficiently large. Robustness is defined as the capacity of agent to recover from failure. Winikoff and Cranefield’s (2014) analysis ratify the BDI failure recovery mechanism as effective as it facilitates achievement of low proportion of real failure (< 1%), even in circumstances in which individual actions have reasonably high likelihoods of failure (5%).
From the foregoing reviews, Agent-Based Systems have been executed as a panacea to the service provisioning problems in the 4G mobile networks. Multi-agent system, derived from dovetailing of intelligent and mobile agents, has aided computerization of services and configuration processes in the implementation of service in 4G mobile networks environment. In this research, an extension of the multi-agent architecture has been deployed in the computerization of enterprise-focused service provisioning process in the 4G mobile network. Research findings confirm that the employed method permits operators to automatically abstract a catalogue of potential users during the launch of new services. It is recommended that this research is escalated to include context-aware provisioning of Integrated Management System (IMS) –enabled ubiquitous services.
Balakrishnan, S. (2019) An Overview of Agent Based Intelligent Systems and Its Tools. CSI Communications magazine, Volume, (42), 15-17.
Dumic, G., Podobnik, V., Jezic, G., Trzec, K., & Petric, A. (2007, June). An agent-based optimization of service fulfillment in next-generation telecommunication systems. In 2007 9th International Conference on Telecommunications (pp. 57-64). IEEE.
Fisher, M. (1994, August). Representing and executing agent-based systems. In International Workshop on Agent Theories, Architectures, and Languages (pp. 307-323). Springer, Berlin, Heidelberg.
Kusek, M., Jezic, G., Ljubi, I., Mlinaric, K., Lovrek, I., Desic, S., … & Huljenic, D. (2003, June). Mobile agent based software operation and maintenance. In Proceedings of the 7th International Conference on Telecommunications, 2003. ConTEL 2003. (Vol. 2, pp. 601-608). IEE
Liu, T. H., & Barber, K. S. (1998, July). Selection of conflict resolution strategies in dynamically organized sensible agent-based systems. In AAAI/IAAI (p. 1193).
Luger, G. F. (2008). Artificial intelligence: structures and strategies for complex problem solving. Pearson education.
Monostori, L., Váncza, J., & Kumara, S. R. (2006). Agent-based systems for manufacturing. CIRP annals, 55(2), 697-720.
Petric, A., Ljubi, I., Trzec, K., Jezic, G., Kusek, M., Podobnik, V., & Jurasovic, K. (2007). An agent based system for business-driven service provisioning. In Proceedings of the 22nd International Conference on Artificial Intelligence.
Stewart, I. A. (2009). The complexity of achievement and maintenance problems in agent-based systems. Artificial intelligence, 146(2), 175-191.
Vasirani, M., & Ossowski, S. (2012). A market-inspired approach for intersection management in urban road traffic networks. Journal of Artificial Intelligence Research, 43, 621-659.
Winikoff, M., & Cranefield, S. (2014). On the testability of BDI agent systems. Journal of Artificial Intelligence Research, 51, 71-131.