Management of the Access Network and Service Provisioning

April 19, 1999
Hursti, Jani
Helsinki University of Technology
Seminar in Internetworking
<jani.hursti@hut.fi>

Abstract
This seminar paper analyzes the Internet access service providers, their business and operations. It presents requirements imposed on the service providers and outlines the ways in which these requirements may change as changes in access technology and the market occur. Finally, the paper presents methods and tools to help the providers manage their operations.

1 Introduction *

1.1 Background *

1.2 Research Problem *

1.3 Objectives of the paper *

1.4 Scope *

2 Service Provisioning and Its Management *

3 Network Access Providers and Their Services *

3.1 Measures of Service Quality *

3.2 New Players in Access Provisioning *

3.3 Service Management as a Competitive Tool *

4 Access Network Technology Challenges *

5 Models for Streamlining Service Provisioning Processes *

5.1 Process Improvement *

5.2 Cross Corporate Arrangements *

5.3 Cost Management *

6 Tools and Services for Service Management *

6.1 Business Management *

6.2 Service Management *

6.3 Operations Management *

6.4 Element Management *

7 Future Directions *

Terms and abbreviations

ABC Activity Based Costing
ADSL Asynchronous Digital Subsriber Line
BPR Business Process Re-engineering
GSM Global System for Mobile Communications
GPRS General Packet Radio Service
IP Internet Protocol
ISDN Integrated Services Digital Network
ISP Internet Service Provider
ITU-T International Telecommunications Union, Telecommunications Standardization Sector
TMN Telecommunications Management Network
SMS Short Message Service
SNMP Simple Network management Protocol
UMTS Universal Mobile Telecommunications System

1 Introduction

1.1 Background

"And one and two and -
Cyberspace slid into existence from two cardinal points.
Smooth, he thought, but not smooth enough. Have to work on it…"
-- William Gibson, Neuromancer

The user access to the communications network, be it the Internet or any other network, is perhaps the most critical part of a communications service. After all, user access deals with the famous "last mile", the final connection between the service provider and the actual customer. This is the point where most bottlenecks in data flow can be found, the place where carriers fiercely fight for access to the cable connecting them to the customers, and the place where the actual paying party finally decides whether to use the service or not.

The access network is the tool for network operators, carriers, service providers, retailers, portals, etc. for offering their services to the end-users, be it private individuals or different organizations. The task is conceptually simple: offer as compelling and high quality service as possible with the lowest possible price. But that is really where the simplicity ends.

This paper describes the management problems of two types of operations, access providers and Internet Service Providers. For the purposes of this paper, access providers means those who operate a physical connection to a user of a data network. The connection can be either wireline or wireless. Internet Service Providers sit at the other end of that connection and offer a further connection to the global Internet with a combination of own cabling and rented connections.

To offer a high quality service, just the right thing to just the right audience, and also make some profit along the way, the access provider needs to take into account all the company internal and external issues related to the business. These issues include at least the following:

Technology. In a high moving industry like Internet access, new, compelling, and at the same time demanding technologies are emerging at a rapid pace and are radically changing the services and the way they are managed.

Politics and law. Telecommunications is a partly regulated business. The access to the customers is controlled by all kinds of officials ranging from consumer protection to regulators of competition and ministries of communication.

Economics. The service provider not only needs money from operations or loans from monetary markets to pay salaries or to finance investments. The economic situation also greatly affects the customers’ ability to pay for the services.

Social issues. Social issues concern the characteristics of the market. What are the demographics of the market? Is access service being used by the rich people, poor people, children, men, or women? For what purposes?

Organizational issues. What kinds of people are needed to offer the service? How to keep them happy and content to their work?

Competitive environment. The service provider is not alone. If the service does not please the customers they go and find it from someone else very quickly.

To add to the complexity, there is not only a large number of issues to consider but also a wide variety of ways to solve them. To improve the service, the access provider can invest in more access capacity, build cables and wireless base stations, make a better contract with long distance carriers, or perhaps invest in an automated administration system. If a customer complains that a competitor offers better service one could recruit more personnel to improve the service, train the existing staff to do their job better, or outsource some help. Or if the company has enough money, why not buy the whole competitor. So much for comparison and complaints, really. Or would it still be wiser to use all that money to recruit a really good lobbyist? He could make sure that the ministry of communications awards you the permit to offer next generation access technology instead of giving the permit to the competitors?

Naturally, all the issues of service management cannot be covered in a seminar paper. And perhaps that is not even required. Each service provider and its environment is unique and dealing with those issues is the work managers do in the companies every day. However, what is done in this paper to assist that task is to take a look at models that have been designed to explain the business. Those models are bound to reality by looking at what is happening now in service provisioning. Finally, tools and services are presented to help service providers improve their current operations and to prepare for the future.

1.2 Research Problem

The purpose of the research for this seminar paper is to take a closer look at companies offering Internet access, be the service integrated into one company or several companies. It seeks out to find answers to the following questions:

What are the most crucial components of service that current access providers need to take into account in their product offering?

Who offers access service and how the management burden is divided and coordinated between these parties?

How can operations be structured to make the service level as high as possible while minimizing costs?

What kinds of services and products are available for access providers in helping in their tasks?

1.3 Objectives of the paper

The objective of this seminar paper is, through answering the research questions, to deepen the reader’s understanding of service provisioning and the requirements imposed to companies intending to make money with it. It is targeted for those unfamiliar but interested in the operational core of a service provider and aims to explain the complexities of the operations. Further, the paper aims to outline different kinds of methods of how this complexity can be reduced and how the operations of a service provider can be streamlined and improved.

1.4 Scope

Management of the access network covers basically all the aspects of the business operations of an access provider. The purpose is not however to try to outline a service model for an optimal service provider nor a future strategy for an ISP. This paper is not a strategy guide nor is it a technical guide to management technologies such as Simple Network Management Protocol (SNMP). The paper, however, covers issues that may help in running such a business. These issues include e.g. outsourcing services, automating management, and rationalizing and improving efficiency. The type of analysis is top down: first trying to see how the business is developing in general and then seeing what kinds of requirements this imposes to different levels of service provisioning.

The data network covered is the Internet. As the title implies, the scope of the paper is restricted to operations near the access point to the Internet. To enable the service, one needs those offering the initial physical connection, be it cable or wireless, from users to data networks, and those offering the access to the Internet from there. The former companies are typically local telephone companies or wireless network operators, and the latter Internet Service Providers (ISPs). Because these functions could be integrated into one single company or could operate as separate companies, both of them have been included into the discussion with more emphasis on service provisioning and thus the ISPs.

This paper does not discuss possibilities and issues of for instance software and hardware manufacturers, long distance carriers, regulators and other political entities, or content providers. These parties may however benefit from the information and find it useful to understand the operations of the access companies.

2 Service Provisioning and Its Management

Most people relate access service management purely to system administration. However, once one looks at different examples of what people are actually looking for from their access provider, it becomes evident that there is much more than that:

"I don’t like my ISP. Their service seems to always be down for some reason and I rarely get to dial in when I want to."

"I had trouble with my ISDN last night. After I had struggled with it for at least an hour, I just called my ISP at 1 am. And there was a guy who helped me get the connection up in less than 5 minutes."

"We tried to get an ADSL connection to our company. The provider promised that men would come and install the required hardware to our premises as soon as possible. After two weeks of not hearing from anyone we gave up and just got a simple ISDN from another company."

"I tried to buy an Internet connection to my son so that he could do his homework at home. So I called a service provider to buy it. There was this young kid who started asking me about all kinds of things about some things with funny technical computer terms. I could not understand a thing so I let my husband go and buy a connection from the store next door to his work."

"My service provider’s service is very cost competitive. With just 30 marks per month I can use the Internet as much as I want to."

All these illustrate some element of the service that affects the customers’ perception of service quality. Only a fractionof these is such that a person under the title "System Administrator" can really control it. Therefore, for the purposes of this paper, service provisioning has a wider meaning. It covers the all core services offered by an access provider and their management and quality assurance. Service provisioning thus involves tasks like service planning and building, network planning, tariffing, order handling, billing, help desks, tests and simulations, fault monitoring, and repairs.

Simplistically put, service management is successful if

It improves the provided service

It reduces the total costs of the provided service

which can later be seen as lower price for the customer and higher revenue and profit for the service provider.

Unfortunately, reality is by far from that simple, the biggest reason being that those processes that link the tasks are many times mixed within an organization and also across the borders of different partnering companies. Tasks and responsibilities are allocated to people in an ad hoc fashion, especially in the smaller and younger companies. This makes it hard to see the distinction between the tasks and even harder to improve and optimize them. In order to make order into the chaos, several attempts have been made to separate the different service, business and technology level considerations from each other. Many of the attempts have been in the form of industry wide initiatives led by organizations like ITU-T [1]. Only with such initiatives it is possible to change the processes of old telecommunication companies that have been operating with totally different market needs and may be reluctant or incapable of changing their operations to adjust to the new challenges.

The results of the attempts are partly overlapping, partly complementary. These include, for example, the Telecommunications Management Network (TMN) model, the Intelligent Networks (IN) model, and the Telecommunications Information Networking Architecture (TINA) among others [13]. The TMN model is used as the guiding model for this paper. The reason for choosing the TMN as the starting point of this paper is that it is simple enough to be grasped quickly and it seems to be widely recognized among industry analysts and equipment and software vendors. Figure 1 shows the TMN model.

Figure 1 Telecommunications Management Network (TMN) model [1]

In the context of the TMN model, business management looks at service provisioning from the top level of the business. Business management is mostly concerned about how to make the service provisioning a profitable business. It looks at the issues like who are the customers to concentrate on, who to employ to take care of the service, with whom should the company have alliances, how should the company compete with the rivals, etc.

Service management can be divided into two broad areas of interest: customer care, and creating and managing new services through the processes of the company. Service management deals with issues like what the value added services are, what kinds of service guarantees can be given to the customers and how those promises are met, how the costs of offering those services can be reduced, how customer requests are answered, etc.

Network and systems management resembles network management done in all organizations with computer networks. Also the tools and processes are fairly similar. Here the tasks fall into the categories performance management, configuration management, accounting management, fault management, and security management. It involves issues like can the network take the load of the services, are all the servers up and running all the time, how bandwidth is allocated to different services, are all the computer systems safe and backed up, etc.

Last but not least, element management looks at the individual network devices themselves. Network elements are typically controlled to find out things like what is the load of individual routers or switches, how many packets flow though individual routers, are the modems correctly configured and running, etc.

At first, the model may seem self evident or even artificial. However, in reality attaining even this much clarity in a real company is a big task. The business moves forward so fast that companies find it difficult, and even dangerous for their flexibility, to try to organize along the lines of a simple model. If the original processes are intertwined, it may seem to be an impossible task to separate and recognize them. Though the reality may be different within a company, the value of the model becomes evident from the management point of view once one sees that some parts of the pyramid can be shared between companies, outsourced, or automated. This paper uses the TMN framework in a top-down fashion. First chapters 3 and 4 look at service quality as a business challenge. Next, chapter 5 shows how service management operations can be improved. Finally, chapter 6 describes different tools which, though positioned to all levels of the TMN model, are predominantly network and systems management and element management tools.

3 Network Access Providers and Their Services

3.1 Measures of Service Quality

It all started with the military and universities offering their personnel access to a packet based network. Now, it seems like everyone would like to offer Internet access services. A wide range of service providers operate especially in the consumer market. In the corporate market the variety is smaller, perhaps because of the high management needs and service quality demanded by the companies. What is it then that the customers are looking for from the service provider? Table 1 shows one set of qualitative measures.

Table 1 Qualitative Requirements for Service Provisioning (Compilation from different sources [1, 11, 8, 12, 9, 10])

Total cost of ownership

Order and purchase processes

Speed, accuracy and quality of installations

Throughput and transmission quality, mobility

Friendliness and professionalism of customer support and care

Accuracy and format of billing

Speed and quality of repairs

Reliability of the service and uptime of the servers and the network

Note that total cost of ownership is listed instead of price. Especially the corporate customers analyze cost sources beyond price. Such include investment in new equipment like modems, computers, possibly even new furniture, the need to employ own people to manage systems, the risk associated with getting locked-in to a service provider, etc.

A fairly new phenomenon, happening commonly in the United States [11] and gaining popularity also in Finland [10] is service level guarantees that are given by the ISPs to guarantee the listed characteristics. A service level guarantee could e.g. say that the network is available for 99% of the time, application availability is 95%, or state guarantees for technical measures like latency, packet loss, etc. These service level agreements are especially important for companies like e-commerce shops who count their operations on the availability of the service [9].

In addition to the qualitative measures that set the baseline of the service, the service providers are offering value added services to the users to attract and retain new customers. Table 2 shows examples of these services. As one can see, the actual access method (modem, ISDN, wireless) is not mentioned anywhere and is not relevant either. The method that best delivers the value offered is the preferred access method.

Table 2 Value Added Applications and Services (compilation of different sources, including [13, 11, 7], and various ISP web sites)

Web hosting	Web hosting means that the service provider offers disk space on its web server to be used for the web pages of the users
E-mail	Most service providers offer an e-mail account for their users either for free or on an additional monthly charge
SMS messaging	In SMS messaging the service provider relays e-mail messages to cellular phones (and vice versa) in Short Message Service (SMS) format
Portal	A portal service means that the service provider has a "starting point" for the World Wide Web users, a collection of information services and links to act as an easy way to start "surfing" the Web
Domain name registration	Typically combined with e-mail and web hosting services, the service provider can register an individual domain name for the customer
News	Usenet news may be available from a news server run by the service provider
Chat	In chat the service provider maintains a chat server to where users can log-in and discuss with each other via a text interface
Managed Firewalls	Some service providers offer to protect the networks of their customers by managing a firewall complying to a set security policy
Virtual Private Networking	Virtual Private Networking is a term used in many different contexts but here it means that the service provider establishes a secured, private tunnel over a public network between two or more premises of the customer
IP Telephony	In IP telephony the customer can make telephone calls that are carried over the Internet instead of telephony networks resulting in low costs of the calls

Note that all services additional to just a basic connection are listed here as value added services. Some may however argue that many of them like an e-mail address or Usenet news are so ubiquitously available from all service providers that they are already taken for granted and no longer constitute anything "value adding".

3.2 New Players in Access Provisioning

After seeing what the access providers offer, let’s take a look at what kinds of organizations offer such service. One can see the boom of the Internet Access provisioning for instance by looking at such markets as the United States. For example, one can find almost 500 different ISPs servicing the area code of California alone [2].

The reason of having so many alternative service providers is partly an issue of service management. If the management of the actual equipment or wiring is fairly straightforward and if regulation in the country gives political assurance that the cables will be available for service companies to use at a competitive price, the entry barrier to start a new company is fairly low. Some ISPs already refuse from reselling access equipment and cabling at all because it is already a "commodity" [11]. Further, new operator technology is removing the burden of the ISPs having to terminate the calls with their own modems. In the new systems, the modems actually reside at the teleoperator from where the ISP can rent ports and get IP traffic to its premises through a single pipe [12]. This further decreases investment needs and transfers ISPs from being "people who have a modem you can call to" to "telephone companies billing and sales partners".

But the access provisioning is not restricted to just teleoperators or dedicated ISPs. For some companies the fact of having some existing infrastructure connecting them to users enables them to also provide wiring and connections to the consumers. This is the case of, for example, water, gas and power companies who now build data cables alongside their other "tubes". Electricity companies are hungry to see whether the power lines themselves could be used for data transport.

Especially in the phase when the market is growing one of the biggest tasks is to reach an unserved clientele. This is the reason for the market entry of those who "own" an existing customer base. This is true for electricity, cable and gas companies but also for many others. Magazines are offering their readers Internet access. In Britain retail chain stores offer Internet access for free. Most people in Germany get their first access service (and can even rent a modem) from their bank. In Spain a parcel service has leased connections from Telefonica and is now offering access through their local logistics service centers. Universities are increasingly wired and offering access to students.

3.3 Service Management as a Competitive Tool

Competition in the access provisioning arena is ruthless. Though cabling can be leased and managed easily, those who own their own access technology can get their revenue from cable usage and press the price for the service to a minimum, straining those who don'’t own their own data network. This may happen even if telecommunications regulation tries to impose a fair pricing policy to all.

Under the competitive pressures it seems like access providers are resorting to two alternative directions. The first direction is a high quality, differentiation choice. Here the company hopes to acquire loyal customers and exclusive experience in managing a certain customer base. The company can then offer these customers exclusive service and content and hope they won’t switch to a lower price provider. Additional money can be perhaps gained from advertising revenue, consultation services, etc. The second choice is cost competition. It requires the pruning of all unnecessary activities, streamlining everything, and possibly making the whole service self service. This choice also may require the building of an alternative access network, through air or ground. And off you go to a price war.

Either way, service management plays a big role in making the choice successful. Especially for those who offer access but do no’t own the cables and d no’t have a reputation of good customer service it means just one thing: time to start to learn access service management and service streamlining. This is an absolute necessity so that they can cut all the costs of basic service and improve the quality of value added.

The size if the service provider and breadth of services gives advantages. In many cases the customer prefers getting the whole package - access equipment, cabling, IP connections and service - from a single vendor and have all those charged on one bill [12, 10]. Smaller players can only count on partnerships. One party takes care of access equipment, the other the Internet connection and collecting data about connection time, the third (e.g. a magazine) takes care of the billing and customer administration, the fourth takes care of branding the service.

From a business partnering perspective forming a partnership is an issue of whether all the partners can be trusted and controlled properly. If one decides to break the agreement or goes bankrupt, the rest are left without a valuable partner. This of course very much affects the reliability of the service. It is hard to offer a certain service level to the end customers if the partners are reluctant to promise any service level to the access service provider. Vice versa, this kind of arrangement can improve the reliability by giving the chance to create redundant relationships. An example if this is an ISP making a deal with multiple long distance carriers for offering access to backbones.

From the service management point of view, the issue is then really whether such a coalition can be streamlined properly. An example could be that one party uses one type of network management software and would like to have that data available for the network planning or tariffing applications of the others. Still, technical compatibility may prove easy compared to contracts and coordinating the personal linkages between the companies.

4 Access Network Technology Challenges

The rapid advancement of technology is enabling new kinds of methods for end users to access information networks. The old analog modem is challenged by technologies running on copper, fiber, and coaxial networks or transmitted from base stations and satellites over the air. Table 3 summarizes a few of these and some of their main characteristics.

Table 3 Available Access Technologies (compilation from [5, 7, 3])

Type	Data rate	Transmission medium	Terminals	Price of end user equipment
Analog modems	14.4 Kbps to 56 Kbps	copper lines	External boxes with serial connection, internal cards, PC cards.	Less than $30 - $500
ISDN	2 x 64 Kbps, 30 x 64 Kbps	copper lines	External modems, internal cards, PC cards, routers	$150 - $500
xDSL	Upstream 64 kBps to 1.5 Mbps Downstream 1 Mbps to 9 Mbps	copper lines	internal and external modems	$250 - $1,000
Cable modems	Upstream 128 Kbps to 10 Mbps Downstream 10 Mbps to 27 Mbps	Hybrid fiber/coax	External box	$300 - $900
1st and 2nd generation mobile wireless access (e.g. GSM, GPRS)	9.6 Kbps to 100 Kbps	Air, radio frequencies	Data modems (internel/external), access through mobile device itself	$300 - $1000
3rd generation mobile wireless access (UMTS)	2 Mbps	Air, radio frequencies	Data modems (internel/external), access through mobile device$	$600 - $1000
Satellite	Upstream 56 Kbps Downstream 400 Kbps (2 Mbps/16 Mbps by 2003)	Air, several frequencies (Ka-band, Ku-band)	Mobile receivers, stationary disks	Under $1000
Microwave links, radio links	10 Mbps - 100 Mbps	Air, e.g. 400 - 900 MHz, 15 GHz	Stationary antennas, networking cards	$3000- $15000
Optical beam transmission	155 Mbps	Air, laser	Stationary receivers	$5000 - $20000

The impact of new access technologies on service provisioning and service management depends a lot on the case and the level of analysis. They could open new opportunities or threaten seriously the current business. The amount of trouble largely depends on the access technology and the type of services offered. For instance, if the service provider does not operate the access network itself, it does not make much of a difference whether the user dials in with a stationary fixed phone, ISDN, or a GSM phone with a data modem (except if e.g. the service provider gets complaints about its web site being so full of graphics that it takes a while to download it at 9.6 Kbps).

However, if the access provider operates the access network itself, the change may mean a huge investment into cables, base stations, or both. Adding a new access technology to the portfolio of products could require at least: buying new termination equipment, making changes to the carrier network, training personnel to use new equipment, updating all network management systems to reflect added equipment, designing new applications that are enabled by the access technology, rethinking pricing and establishing new pricing for the new service, making necessary changes to user databases and billing systems, testing and simulating all applications under different scenarios before service launch, etc.

Business management needs to carefully look at how the technologies change the market dynamics. For instance, in the United States cable companies are threatening incumbent operators especially in the consumer market. Cable operators have existing wiring and customer relationships, can offer very fast connections, and are used to offering entertainment to homes, something that is expected to be the main thing home users are looking for on the Web. As another example, in Sweden tests are being made with radio links to avoid using the cables of the incumbent operator Telia which owns most of the networks in Sweden.

5 Models for Streamlining Service Provisioning Processes

In the digital age, the efficiency to serve a large customer base is many times sought through automation. But one has to notice that the reason for not being able to scale the service might not be because the machines do not scale or cannot handle a large customer base but simply because the employees cannot handle such a customer volume. There may not be enough people to resolve problems or help the customers. Or simply there is just no-one to process all the orders that have been faxed to the office. Or even if the service provider itself is able to resolve the problem in its own equipment, the partners in the chain can just say "no can do". Solving these problems s the task of process improvement.

5.1 Process Improvement

Streamlining and rationalizing the service tasks performed by employees is of crucial importance when serving a large volume of customers. This can be done with plain common sense or by following more formal models such as Business Process Re-engineering, BPR. In BPR the tasks in a process are drawn as accurately as possible and rationalized by reducing cycles between tasks and pruning unnecessary steps. Without going into specific details of BPR, let’s take an example of a customer order process which might look like the one in Figure 2.

Figure 2 An Example of a Possible Order Process at an ISP Before Process Re-Engineering [14]

No automated system can help to reduce redundancy without first ordering the tasks differently. However, hopefully after a successful process reengineering the process could look like the one in Figure 3. It is now much easier to automate the necessary tasks and automation might actually prove unnecessary if the process already runs smoothly enough.

Figure 3 An Example of a Rationalized Order Process at an ISP[14]

Let’s see what has actually happened in the process. First, the aim is to break down the functional barriers and have people understand what other parts of the organization do. This way they can "serve" other parts accordingly. For instance, salesmen cannot "just sell", they have to actively function as the interface to the customer and to provide sufficient information to customer care people. Second, by increasing the skill level of the workers, their priorities and quality of their work all kinds of "notifications" can be eliminated. There is no need to check whether a service is up and running if one can trust that the service managers do their jobs properly.

The end result is that the customer gets the service faster and more reliably, workers are better motivated as they know their contribution to the whole better, and less management is needed to coordinate the tasks.

5.2 Cross Corporate Arrangements

The real challenge for process improvement is extending the improvements to business partners to reduce bottlenecks. By looking at the different kinds of arrangements that exist between networking companies one can see the difficulty. Examples of such arrangements include:

Leasing contracts where service providers lease cables from incumbent local telephone operators

Contracts for system integrators to install cabling and equipment on behalf of the service provider

Contracts between local service providers and backbone operators

Service contracts where a third party does network management for the service provider

Billing contracts where a third party takes care of the billing

Roaming deals between ISPs where a user gets access to its own operator by calling a partner operator

Peering deals between ISPs where two or more ISPs agree to exchange traffic at private access points and can thus avoid congestion at public access points

etc.

The trouble is that many ISPs operate with fairly vague arms-length agreements with their partners. The processes do not match, the systems do not match. The problem in trying to improve the processes is that the operators are not eager to tell the partners, who might simultaneously be their competitors, how they have arranged their systems. Internal systems are sources of competitive advantage which are not easily revealed to outside organizations.

To remedy the situation the industry has established programs like SMART (Service Management Automation and Reengineering Team) that have established a set of inputs and outputs from each task to another (like a public interface between companies). By fitting the internal tasks to the external interfaces the companies should have better chances for successful cooperation [1].

5.3 Cost Management

At the same time the process is improved, the service provider must also look at the costs. Several cost control methods are available. The most famous is perhaps Activity Based Costing, ABC. Though ABC was originally developed for manufacturing operations, many of its principles apply to service also. The purpose of Activity Based Costing is to minimize the use of cost allocation with general "overhead costs" and try to analyze in detail how much each activity really costs. ABC allocates costs to so called cost functions. Through the allocation one can really see what specific tasks like sending a bill, making a repair visit, or answering a support call really cost. This information can then be used for accurate service pricing.

6 Tools and Services for Service Management

Started from the technology, the access service business seems to be one of the most prominent businesses for process and task automation. After all, the task is to get people connected with machines to other machines and it would sound natural to have even more machines to control the whole thing, right? Seriously, offering access to masses of information to masses of people makes access service provisioning very much a volume business. To control all that mass the processes necessarily have to be automated to ensure the smooth flow of information at the set service level.

The tools that are offered to ISPs for service management very much rely on different kinds of statistical information that can be collected from the networking devices. There seems to be so much data available about different events that it should be enough for anyone. The actual decisions and management of processes are still left to the human managers. Although automation is available, there is still not a device that would act as a crystal ball an predict how those applications will succeed or whether the users really need the latest access technology. Though the names of the products may imply otherwise, you cannot buy a guru in a box. There is still plenty of room for the famous managerial "judgment", "hindsight", and "vision".

Many of the presented tools are standalone systems from fairly small manufacturers. Large companies like IBM (www.ibm.com) and Hewlett-Packard (www.hp.com/go/telecom) have also awakened to address the market and with their offerings an ISP can now build an integrated system covering basically most of the needs of service provisioning from element management to customer databases, billing, and help desks.

Where automated tools are not available, there is a plethora of service companies to assist the service provider in trouble. Management consultants can help making right business management decisions. Service consultants can do process reengineering or test applications and do pilots. Specialized network management companies can take care of installation and monitoring of servers and software. Billing and help desks can be outsourced as well as marketing. Actually, if required, the service provider can really be a fairly hollow company.

The analysts of the IT networking business such as Gartner Group are using the TNM model extensively in tasks like grouping the available software into categories based on which task in the TMN model those tools are meant for. The categorization in the following description is based on the original categorization by Gartner Group [6]. However, as can be seen, the differences between the functionality offered by the software does not differ so clearly that the categorization is really realistic. It is however retained in this paper for clarification and convenience and for the reader to compare whether either Gartner’s classification or the positioning of the vendors is correct.

6.1 Business Management

Based on the classification by Gartner Group, products that would fall into business management category include such products as Chariot and Pegasus from Ganymede Software Inc. (www.ganymede.com), Application Expert by Optimal Networks (www.optimal.com) and ITDecisionGuru from Third Millenium Technologies (www.mil3.com). These products are mainly meant for proactively testing and simulating different usage scenarios of a network. Such situation could e.g. be testing whether offering a Voice-over-IP service requires investment into new networking equipment as the bandwidth usage increases.

Basically these tools provide more and more statistics based on either historical data or a simulation of the future. Most of the tools are not something the CEO of a large company would utilize in his or her daily work. A more probable user group is technical managers and middle management. Though fairly sophisticated tools are available, the possibly most used tool for actual business management is the Microsoft Excel spreadsheet program.

6.2 Service Management

Service management tools help the service managers ensure that services run as promised. One of the keywords here is service level agreements. Products such as Proviso from Quallaby Inc. (www.quallaby.com), S3 from NextPoint (www.nextpoint.com), and EcoSYSTEMS from Compuware (www.compuware.com) help the service managers monitor the network throughput and application usage. The aim is to be able to proactively test what the network and the applications can take so that the performance can be guaranteed to the customers. Of course, continuous monitoring is needed to spot the possible problems before the customers do.

Most of the tools seem to concentrate on displaying different kinds of statistical data very similar to the data available from products meant for operations management. Such data is typically technical in nature, e.g., packets lost, packets available, congestion, collisions, overall load and availability, etc. Even "executive level" charts are about latency, bandwidth utilization, daily/monthly/weekly network volume, etc. Perhaps the difference to systems meant for network monitoring is the ability to automatically analyze different kinds of trends for the future in the service.

Though applications can be tested before actual launch to debug errors, only specific consultation services test the actual performance and application availability in qualitative, not quantitative format. Consultant services can go much deeper into the service by e.g. analyzing the actual physical premises and wiring and by giving their personal opinions about the services.

Billing and customer databases form an integral part of service management. In most cases, these customer care processes can be handled by traditional database and invoicing applications that have existed in companies for decades. Especially when the pricing models are based on so called cyclic billing, which means regular, flat fee bills this can be done with the same systems as e.g. the ones used for magazine subscription billing.

Billing software dedicated to ISPs seems to be very much in its infancy. Billing at ISPs can be enhanced by taking into account specific data that can be collected from e.g. RADIUS servers or the Microsoft Commercial Internet System. This kind of data gives more freedom to design new tariffs and billing models. Examples of systems targeted for ISP use are Eclipse from SunTech Systems (www.suntech.com) and Internet Management System from Belle Systems (www.belle.dk) (distributed by Hewlett-Packard). New technologies such as certificate based authentication will also open new possibilities for more accurate and faster "on-the-fly" charging for the services. As companies build business models based on different cooperative agreements, the billing systems will have to accommodate this. An example of such a model and system is the iPass (www.ipass.com) that enables roaming between ISPs. As the user roams through the networks of different ISPs, the fees for this use are taken from the "home ISP" of the user.

Additional service management tools include automated help desks or tools that help help desk operators gather and disseminate information quickly. These kinds of systems not only include internal Frequently Asked Questions databases and response tracking but also customer history and other data.

6.3 Operations Management

Operations management is an area where such systems as OpenView from Hewlett Packard (openview.hp.com), Tivoli from IBM (www.tivoli.com), and CA Unicenter TNG from Computer Associates (www.cai.com) help network administrators keep their systems up and running. This is an area of "traditional" network management where the performance, configurations and errors are tracked. These applications have grown from simple SNMP management stations to large scale monitoring systems which may include also features like help desk or usage accounting. Perhaps the real difference with this category and service management is that these products are targeted for "general network management", not just service management. There are though versions of these products available that have been tailor-made to service providers.

6.4 Element Management

Whereas tools for business, service, and operations management are made by vendors dedicated to the management tasks, element management systems are typically manufactured by the network device vendors themselves. These are systems that are used to monitor and control individual devices. Their functionality is typically dedicated to the vendors own devices, though they might be based on standards like SNMP and might even be built on systems like OpenView. They can automatically discover devices on the network and let administrators configure and monitor them.

7 Future Directions

The problem with consumer access service is that it is like the entrance fee to a large amusement park. Like in the case of the amusement park entry, the visitors are reluctant to pay for the entry because they know that they still have to pay additional payments for the rides and games. They are there for the actual rides so why charge for the entry - it does not provide any value. True, on the Internet there are free services available for the access fee only and its certainly fun to just go to an amusement park and payment for that is justifiable, but it still seems awkward. And why should I pay the same amount as the guy next door if I only take one ride and he stays there for the whole day?

The difference between Internet access and amusement parks is that there are multiple entrances run by different companies. Consumers have a choice and an option to complain about awkwardness. The awkwardness of entry pricing can be seen in competitive Internet access markets where there really does not seem to be a reasonable price for the access. It could be $100/month, $10/month or totally free. "Access" is just too vague a term to attach any price tag to it.

Its an interesting dilemma. Without the entrance the services could not exist but no-one wants to pay for the entrance. No wonder the access providers are now looking for alternative sources of revenue. To use the amusement park analogy they could

charge road tolls for the roads to the parks (operate own access cables);

take rents from the people operating the rides and games (offer access to companies who use the Internet for business);

rent or sell bikes and shoes for people to move around in the park (sell access equipment to end users);

sell pop-corn at the entrance (e-commerce);

offer guides and maps as a service for park visitors (portals);

get an external sponsor at their entrance (web advertising); or

charge a percentage of the revenues of ride operators (team up with service companies)

At the same time the industry matures and smaller players drop out. Companies and correspondingly their systems and organizations get larger and more complex. As penetration increases the management operations move from improving efficiency in order to reach more and more customers to retaining the customer base and reducing churn.

The alternatives seem to imply that a large part of the service management will actually concern managing a bundle of contacts to other companies and presenting these in a processed format, be it called a portal or whatever. Still, the providers cannot forget the quality of the actual access. The times when America Online got a lot of negative publicity because the service could not serve everyone are not far behind. It is a challenge and a battle where only the fittest - the best managed - survive.

References

[1]: Adams, Elizabeth K.; Willetts, Keith J. The Lean Communications Provider. McGraw-Hill, New York, 1996.
[2]: The CMP Media ISP Directory ISPs.com. [referred 28.3.1999] <http://www.isps.com>
[3]: Data Communications (CMP Media). Global Enterprise Networking Directory. [referred 18.4.1999] <http://www.data.com/Directory/>
[4]: Gibson, William. Neuromancer. Ace Books, New York, 1984.
[5]: Girard, J.; Payne, S.; Pultz, J.; Hafner, R.; M.; Smith, C.; Egan, R. Remote Network Access Technologies: Addressing the Need for Higher Speeds. The Gartner Group, 1998
[6]: Govekar, M.; Gassman, B. Network Management Tools Segmentation., The Gartner Group, 1998.
[7]: Heywood, Peter; Dornan Andrew. Looking to Go the Last Mile. Data Communications International. December 1998, pp 40A-40D. CMP Media Inc., New York.
[8]: Heywood, Peter; To Chee Eng. International Service Providers - The Best in the World. Data Communications International, May 1998, pp. 72A-72I. CMP Media Inc., New York.
[9]: Hulkkonen, Juha, Project Manager, Hewlett Packard Oy. Interview, Espoo 8.4.1999.
[10]: Kuusela, Jyri, Product Manager, Sonera Oyj. Interview, Helsinki 14.4.1999.
[11]: Makris, Joanna. Guerilla ISPs. Data Communications International, May 1998, pp. 90-102. CMP Media Inc., New York.
[12]: Rehnberg, Petri, Technical Director, Netti Finland Oy. Interview, Turku 7.4.1999.
[13]: Tuisku, Miika. Service Management in an Online Service. Master’s Thesis, Helsinki University of Technology, Department of Computer Science, Espoo, 1998.

Notes

[14]: Source: the author. The figure is an illustrative example of a possible sales and service startup process within an access service provider. Though the levels resemble the ones found in real life organizations no particular real life process has been used as the basis.