AutomatedBuildings.com Article - WEB ENABLED GATEWAYS Part 2: Connecting a Gateway to the Software Application

This series of articles will explain gateway features and provide a basic understanding of what is important in the "alphabet salad" above.

In the last article, I listed some questions about information transfer requirements and the available site's infrastructure. The answers to these questions are crucial for successful gateway project implementation and low operations costs. This article explains the questions and the ramifications of the answers.

The questions asked in the previous article were:
Is there a corporate network, and if so, are there connection points available at the planned server and gateway locations?
Are there limitations on network use?
Are there times where bandwidth is unavailable?
Are static (fixed) IP addresses available or is only dynamic (DHCP) IP addressing available?
Is there a cost in using the network, and if so, how is it priced?
Are there restrictions on the TCP/IP ports and protocols that may be used?

If there is no corporate network, do the sites have an Internet connection via Broadband or DSL?
Is there an existing telephone line that can be shared?

Information transfer requirements have a direct bearing on the communications infrastructure required.

Typically in the past, the communications infrastructure was a telephone line and a modem, with the gateway being polled periodically by a server, or dialing out and reporting data periodically.

Some systems used wireless rather than telephone lines but the basic concept is similar apart from the difference in cost and the more limited bandwidth of most wireless systems.

This form of information transfer can be classified as intermittent communications, which is server or gateway initiated. This type of gateway is also commonly called a data logger or remote terminal unit (RTU) with data logging capabilities. Intermittent communications inherently assumes gateway data logging.

Intermittent communications along with associated data logging is suitable for reporting applications where the latency, or periods between data transfer, is not critical. Typically, the logged data is stored by the application and is used to provide reports.

Limitations of this type of system are the cost of modems, telephone line rental and call charges. Call charges may also limit the frequency of data transfer and the volume of data transferred. Data storage size in the gateway also determines the number of points, data logging rate and maximum time between uploads.

Where direct modem server-to-gateway communication is used, managing the uploads, sharing telephone lines and managing costs limits scalability and the amount of data gathered.

This form of data transfer is not suitable for alarming systems due to the extended latency between uploads. Some systems solve this problem by having the gateway "push" alarms to a server independent of the data logging uploads. This requires gateway initiated communication and a server that can manage multiple incoming calls.

Low cost dial-in ISP services have lowered communications costs for those gateways that can communicate via standard Internet protocols such as PPP. Instead of a direct server-to-gateway modem connection, the gateways dial into an ISP via a local telephone number and communicate via the Internet to the server. As the server does not require a bank of modems and calls are local, costs can be reduced.

Unfortunately many modem-based gateways use proprietary communications protocols that are not Internet enabled. This makes these systems impossible to upgrade to take advantage of Internet technology.

Gateways that do support Internet enabled modem communications are a viable solution for sites where there is no LAN/WAN or Internet connection via DSL, Broadband, Satellite or other media.

In general, modem based communication can be used for periodic uploads and limited alarming, but is not viable for a continuous connection due to the cost of continuous use of a telephone line, or the cost of a leased line.

Corporate Networks
Due to the communications boom in the last few years, many buildings are wired for LAN (local area networks) for corporate business use. These LANs are often connected to the Internet and interconnected to form corporate Intranets, or WANs (Wide area networks).

To use a LAN or WAN, the gateway must have an Ethernet port (typically a 10BaseT connection) that MUST also support a network protocol such as TCP/IP. There are many systems that use Ethernet, but due to proprietary or uncommon protocols, cannot communicate via WAN or the Internet.

Intranet or Internet connections can provide a low or no cost information transfer infrastructure. As the special communications needs of gateways, and the perceived threat to network security can affect the viability of this option, careful matching of gateway capabilities and corporate network policies is required. (More on matching gateway capabilities and networking policies later in this article.)

Dependent on the bandwidth available, a connection to a LAN can enable a continuous connection that provides added benefits, in terms of semi real-time monitoring as well as reduced gateway data storage requirements.

Dedicated Connections
Dedicated Internet connections are now commonly available in most areas at costs that can make these options feasible. These Internet connections are Broadband over cable, DSL over telephone lines and dedicated satellite.

Broadband and DSL connections require that the service be available at the sites. DSL also has a distance limitation between the local telephone exchange and the site location. As the broadband and DSL markets are served by a variety of vendors that are divided geographically, an application with a large number of geographically scattered sites requires dealing with multiple service providers.

Satellite, which does not have the DSL or broadband limitations, is typically more expensive and has lower bandwidth. The inherent delay in satellite communications that is frustrating for a typical user does not affect gateway communication if the gateway is designed to take these delays into account.

Dedicated connections can be viable for sites where corporate connections are unavailable or the cost or "hassle factor" of dealing with corporate IT makes using a corporate network unattractive.

When considering corporate or dedicated connections, the cost and feasibility of these connections often is influenced by gateway technical capabilities.

Gateway technical capabilities and how they influence operations and implementation costs.

Both the cost of a dedicated connection, as well as compliance with IT department's policies, may be dictated by bandwidth requirements as well as IP address assignment.

IP Address
An IP (Internet Protocol) address is required by every gateway apart from gateways that use direct modem based gateway-to-server communications. The IP address determines the identity of the gateway and is required for LAN/WAN or Internet communications. The IP address can be configured at the gateway or can be assigned by a network DHCP server. A configured IP address is a static address, while an IP address assigned by a DHCP server, or the ISP equivalent, can change at random intervals or when the gateway is powered up or connected to a network. Assigned IP addresses are common on corporate networks as they make the network easier to administrate. In some cases, static IP addresses are available on corporate networks. Static IP addresses on dedicated connections almost always bear a significant additional cost and may be unavailable in some cases. Gateways supporting dynamic IP addresses are required in situations where static IP addresses are unavailable or expensive. Dynamic addressing requires that the gateway initiate communications to the server as the server can never be sure of the gateway's addresses and cannot initiate a conversation.

Bandwidth
Bandwidth is normally specified in terms of bits per second which can be the same (symmetric) or different (asymmetric) for uploading and downloading data. The available bandwidth can vary dependent on whether or not the link is shared, as in the case of broadband. In some situations, the available bandwidth varies according to the time of day. In such cases, having the ability to schedule log file transfers from the gateway to the server can make use of available bandwidth windows.

Security
Security has two aspects: secure data transfer, and whether the security provisions already in place will allow the gateway to communicate with the server. (Data security will be discussed in a later article when I discuss communication with server applications.)

Security provisions and security policies can directly affect the implementation of a gateway system. Many corporate networks lock out communications over any port apart from a few commonly used ports. A port can be thought of as a communication channel that allows devices to communicate via TCP/IP. A common port is port 80 that is used as a default by web servers.

The side listening for communication listens on a particular port. As we saw earlier, if the gateway has a dynamic IP address, it must initiate the communication and the server is therefore the listener. In this case the communications infrastructure must allow communications traffic to the server through the specified port. If the gateway has a static IP address, the server can poll the gateway provided that the communications infrastructure allows communications to the port that the gateway is using.

A gateway, whose communication protocol can use standard ports, and the commonly available protocol standards associated with the ports, has a greater chance of adhering to security policies and provisions.

The next article will discuss communications protocols, server applications and how to interface with server applications.

Events	Want Ads
Our Sponsors	Resources


	Jack Chadowitz, CEO BostonBase Inc., jack@bostonbase.net