Internet Call Center

A call center is a system that allows a group of people (agents) either to make calls to or take calls from customers. The system usually provides the following functions:

§  Queuing of incoming calls. A typical queuing discipline is first-come, first-served, but often large-volume customers are given priority.

§  Call distribution. Incoming calls are typically distributed based on the agents’ availability. With the feature known as automatic call distribution (ACD), any given call is routed to the agent who has not received calls for the longest period of time. With another feature, that call can be routed to the agent who can best deal with the caller’s problem. For best-agent calls, the caller is first connected to the voice response system, which prompts the caller to make a choice from a list of options, and then is transferred to the appropriate agent.

§  Agent monitoring. The system tracks the time each agent spends actually responding to a call, the duration of each call, and other statistics that report on the load and efficiency of agents.

The Internet call center augments the capabilities of a traditional call center with the access to the Internet in general and to the World Wide Web in particular. In some cases, it also enables agents to use IP telephony in addition to the PSTN. In all cases, it gives the agent access to e-mail, text chat, and escorted Web browsing. These services are used to improve the information exchange between the customer and the agent. For example, voice communication may also be supplemented by IP-based information services. With escorted Web browsing, the agent can guide a customer through a set of Web pages, viewing the same pages as the customer.

Multimedia Conferencing and Data Collaboration

Multimedia conferencing services support real-time communications among users at multiple (usually more than three) locations. The communications use a combination of various types of media such as voice, video, text, graphics, and still pictures. Following is a list of common forms of multimedia conferencing services:

§  Audio-graphic conference. Adds nonvideo data exchange to a normal audio conference. Participants of the conference can, for example, view the same presentation slide as the presenter during his or her presentation.

§  Video conference. Uses video and voice for communications for a virtual face-to-face meeting.

§  Data collaboration. Allows the participants not only to share, but also to edit documents (like word processor files or spreadsheets) jointly.

Multimedia conferencing services can be provided over various types of networks. Figure 1 depicts a scenario where audio conferencing takes place over the PSTN while data conferencing takes place over the Internet. Naturally, the conference participants need both voice access to the PSTN (via a telephone terminal) and data access to the Internet (via a PC). This access could be achieved in several ways:

Figure 1: A multimedia conferencing scenario.

§  With two POTS lines

§  With a modem that splits a POTS line into voice and data channels

§  With an ISDN line

§  With an xDSL line 

§  With a cable or power line, or wireless or enterprise LAN connection for the Internet access

Figure 2 depicts another scenario where audio conferencing takes place over the PSTN as well as an IP network, while data and video conferencing take place over just the IP network.

Figure 2: Another multimedia conferencing scenario.

Regardless of network configuration, conference control and management (or operation) of the conference resources (like media and bandwidth) are important aspects of the multimedia conferencing services. Conference control provides the means for a conference to proceed in a certain order, as determined by a designated conference chairman. The chairman of the conference can give the floor to a conference participant to speak, or grant permission to a participant to transmit supporting media. A participant making a presentation naturally should be given the right to speak as well as transmit and control the visual information. The chairman may also mute a speaker, normally for quick intervention with an important message. The chairman may also organize a private subconference (or side conversation) meant to proceed undetected by the excluded participants.

In addition to the conference chairman, conference control also recognizes the role of the conference controller (both roles are often given to the same person, associated with one terminal). The conference controller can perform the following functions:

§  Add or drop a party (terminal)

§  Split the conference or merge a split conference

§  Extend or terminate the conference

For a video conference, the controller may also be able to change the composition of the video screen displaying images from various locations during the conference.

Conference functions are limited by the capabilities of the terminal equipment. Unlike the telephone sets, which have more or less standard capabilities, multimedia terminals are quite dissimilar. For example, one terminal may handle just audio; another terminal audio, video, and fax; and yet another may handle audio and video capabilities, but with inferior quality. The resource management aspect of multimedia conferencing services must accommodate the varying capabilities of different terminals. There are mechanisms for terminals to negotiate capabilities when joining the conference. These capabilities can be renegotiated while the conference is in progress. Following are three common ways to handle differences in terminal capabilities:

1.  All terminals operate in the mode supported by the least capable terminal.

2.  Each terminal operates in its own mode, rejecting or ignoring input beyond its capabilities.

3.  Specialized conference equipment mediates communications so that each terminal receives and transmits only what it can handle. (This mediation may involve transcoding if the media formats supported by the terminals are incompatible.)

In addition to capability mediation, the conference equipment manages resources by performing the following three actions:

1.  Establishing, modifying, and releasing connections to the terminals.

2.  Monitoring the activities over each connection and tracking the status of each terminal.

3.  Handling media streams received and sent through each connection; switching, distributing, multiplexing, and mixing them as necessary.

Note that conferencing service requirements have been studied by the industry for several years and have been standardized by ITU-T. For more details on the generic multimedia conferencing service and features, consult ITU-T Recommendations F.700 and F.702. Collaborative conferencing is an application that can be further subdivided into broad categories in which the same features are used for different ends. Categories include executive meetings (the model for the introduction and discussion of the major conferencing features), distance learning, and telemedicine.

With distance learning, university classrooms can be brought to corporate locations and even homes. It is less important for the professor to see his or her students than the other way around, so video broadcasting is a viable option for distance learning. As the student works, the professor can access and evaluate the assignment and provide feedback in real time. As a result, the student may actually receive more personal attention than a typical classroom student. Organizations that encourage employees to take distance learning classes save both tuition and commuting time.

Telemedicine also includes meetings of doctors and continuing education (required of physicians by law in many countries), but its more ingenious and often life-saving use is in remote diagnosis, consultation, and patient prescreening. With distance diagnosis, the services of the best physicians can be rendered to patients in the remote geographic areas where treatment is not readily available. The transmission of both static (still) images (e.g., X-rays, electrocardiograms, and biopsy images) and dynamic ones (e.g., ultrasound images or patients’ movements) is possible. In special cases where several variously located specialists must all examine a patient virtually, all features of collaborative conferences can be used. Special features include the ability for doctors to make comments unheard by the patient. The savings due to elimination of travel for both doctors and patients is obvious, but the chief advantage of telemedicine is that it can bring health care services of the highest quality to areas where they were previously unavailable.

IP Fax

Fax has traditionally been treated as a telephony application because it is sent over the telephone network. In reality, it is a data application that results in transmitting scanned document images from one facsimile terminal (fax machine) to another. The wide availability of the Internet has created an environment that allows the fax application to return to its true identity as a pure data service. If the real-time delivery of fax is not a requirement (and more often than not it is not), the image of a scanned document can be sent over the Internet in a way that is similar to sending e-mail.

For integrating the PSTN and the Internet, the following three scenarios are relevant:

1.  A traditional fax machine sends a document to another one by way of an IP network. (A configuration is depicted in Figure 1.)

Figure 1: Traditional fax–to–traditional fax.

2.  A traditional fax machine sends a document to an IP fax machine. (See Figure 2.)

Figure 2: Traditional fax–to–IP fax and vice versa.

3.  An IP fax machine sends a document to a traditional fax machine. (See Figure 2.)

Note that an IP fax machine is an appliance, typically a PC, that can be connected directly to an IP network. Because it is necessary to establish a phone call in order to interoperate with the traditional fax machine, the steps involved in making fax calls repeat those already discussed in the IP telephony-call scenarios.

The IP fax operation can take place in two modes: real-time and non-real-time. In the real-time mode, IP fax emulates the behavior of a traditional facsimile over the PSTN. The sequence of events, such as capability negotiation and confirmation of receipt by the endpoints in a traditional facsimile session, are largely preserved.

In the non-real-time mode, IP fax operates much like e-mail, relying on the store-and-forward process. The fax is stored at a staging point before being transmitted to the next staging point. The end-to-end transmission of the facsimile may involve several staging points with an unpredictable delay.

Unified Messaging

Messaging refers to non-real-time communications by way of voice mail, e-mail, fax, and the like. Of these, voice messaging and fax have traditionally been provided by voice networks and e-mail by the Internet. Until recently, each form of messaging required separate storage, access, and management; neither the applications nor the specialized devices worked with each other.

Unified messaging is positioned to eliminate the boundaries across these different forms of messaging. As a result, users can create, send, and retrieve any type of message with any type of terminal anytime, anywhere (Figure 1).

Figure 1: Unified messaging.

(Courtesy of Jackie Orr.)

Users are given access to any type of message from a number of devices that themselves belong to different types of networks. They have the following options for retrieving messages:

§  Receiving notification of incoming messages (regardless of type) on a combination of devices such as telephone, PC, and personal digital assistant (PDA).

§  Using a PC (or PDA) to retrieve voice-mail messages. Depending on the technology and device available, this option provides two further possibilities:

1.     Playing the messages on the PC (or PDA), if it is equipped with a speaker.

2.     Reading the voice messages (converted to text) on the PC (or PDA).

§  Using a telephone to retrieve e-mail messages. The e-mail messages can be delivered to the user as voice-mail messages with the help of the text-to-speech technology. Alternatively, the messages can be delivered via facsimile to a fax machine that the user specifies, after the user listens to the converted message headers.

§  Using a PC (or PDA) to retrieve fax messages, or requesting that the fax messages be delivered to a specific fax machine.

§  Using a telephone to retrieve fax messages and listen to them (here the messages are first converted to text—involving a character recognition application—and then to speech), or requesting that the fax messages be delivered to a specific fax machine.

The user has various options for sending messages as well. You can mix the means for both creating and delivering messages. For example, the user can create a voice message using the telephone over the PSTN and have the voice message delivered to the recipient as a voice attachment (or just as a plaintext message with the help of the voice-to-text technology) via e-mail over the Internet.

Behind unified messaging is the concept of the universal mailbox (see Figure 2), which holds messages of all types in a single logical location. With a universal mailbox, the user no longer needs to be concerned with the location of messages. In addition, the user is given numerous options for handling messages, regardless of their format and means of access. For example, the messages can be:

§  Sorted by using any part of the header information (such as subject, date, sender, length, and priority) as the key.

§  Selected from a list of common operations (such as reply, forward, save, delete, skip, rewind, and fast-forward).

§  Stored in files, which can be further organized into folders.

§  Filtered according to priority, subject, sender, and so on.

§  Mixed with other messages (possibly of various types, so an audio message and a spreadsheet can be combined into one message) and sent to another user.

Figure 2: Universal mailbox. (Courtesy of Jackie Orr.)

Today, unified messaging operations are most effectively performed on the computer. As the limited (and limiting) capabilities of the telephone keypad are being augmented by voice-controlled applications, telephones are on their way to becoming as effective.

A closer look at unified messaging reveals its essence: the use of the e-mail paradigm by voice mail. The universal mailbox is connected to networks of different types and acts as a voice portal for voice-controlled services.

The immediate improvement in services that has resulted from the application of the e-mail paradigm is best illustrated in a relatively new voice messaging application called call sender with rebound. An annoying feature of yesterday’s voice mail was that you could not return a call pertinent to a particular message without leaving the mailbox. You had to either write down on a piece of paper all the numbers while listening to all the messages (and risk forgetting in the process what some of the calls to be returned were about) or keep calling the mailbox back after each call returned. Clearly, this is not how e-mail works; e-mail is more flexible and user friendly. The call sender with rebound feature available in today’s top products works just like e-mail: It allows the recipient of the message to call back and, after having completed the call, to return to retrieving the messages starting just at the point where he or she left off. The procedure of calling back (using traditional telephone) was also adopted from the e-mail paradigm: You can call back by pushing a single button on a telephone keypad or giving a voice command. In addition, the return addresses of voice messages delivered by the PSTN (that is, telephone numbers) can be stored in and retrieved for redial from personal address books. Similarly, the voice confirmation of (voice) message delivery is a very convenient e-mail-like feature, and so are the abilities to send messages to groups and to specify delivery options.

Another interesting application called reply-to-telephone answering enables messaging replies to telephone callers. Through access to an IN database, the contact options of the calling party are captured and presented to the called party, who has a choice of returning a call, sending a voice message to the caller’s voice mailbox, or sending a voice message via a store-and-forward network.

To access voice messaging from the computer, the visual mail application, convenient for small-office/home-office (SOHO) subscribers, provides end users with the ability to see (and edit) complete lists of their incoming or stored voice-mail and fax messages. The messages can be replied to by using the microphone of a multimedia computer; they can be forwarded, copied, and so on, in exactly the same manner as e-mail. Voice messages can be played using specialized player interfaces with controls that allow you to replay specific portions of a message. Callers often leave their numbers only at the end of their messages after they have been warned that only a few seconds of recording time are left—which, in turn, makes them rush when leaving numbers. As a result, the recipient of a message has to listen to it several times in order to write down the number. This visual-mail application comes in two flavors: Web visual mail and PC visual mail. With the former, users access and manage their mailboxes at a Web site using standard browsers; with the latter, subscribers use a standard Internet mail agent to access their mailboxes.

Even without the full set of visual-mail capabilities, the following two simple unified messaging features explore the duality of the vocal versus visual access:

1.  Bridging. Telephone numbers are converted into e-mail addresses and vice versa. Voice-mail users can send their messages to the Internet; e-mail users can send voice messages to the PSTN.

2.  Cross notification. Voice-mail users are notified of their e-mail reception; e-mail users are notified of PSTN calls.

Support of multiple languages is typical for advanced implementations of messaging, but one important language supported by the voice-mail part of the application is designed specifically for the deaf. The telecommunications device for the deaf (TDD) converts tones to alphanumeric characters and vice versa. TDDs have been used for communications with the deaf over PSTN lines, but now they can be naturally integrated with messaging.

As an introductory example of a crude mechanism of establishing something that approximates a telephone conversation (and is available to all owners of multimedia PCs connected to the Internet), consider the following exchange. A person speaks a phrase or two into the PC microphone and saves this message as a file (preferably in a compressed format). That file is then sent via e-mail to another person, who receives it (typically, in a matter of seconds), plays it, and produces a response in exactly the same manner. The conversation can go on like this indefinitely, and, if the files are not too large, this poor man’s IP telephony can be surprisingly efficient. This simple example vividly illustrates the potential of unified messaging.