Today’s high-speed data applications need system integration plus testing that determines performance limits outside of DOCSIS certification procedures.
Extensive high-speed, standards-based, hybrid fiber/coax (HFC) data networks now are a reality, with more than 20 million CableLabs® DOCSIS™ (Data Over Cable Service Interface Specification) modems deployed worldwide. However, both vendors and operators are realizing that implementing these networks using cable modems from multiple vendors is not always a smooth road.
The days of fully tested, integrated, turnkey proprietary systems throughout the network are a thing of the past. The industry has shifted to multivendor, open-system architectures, which inevitably presents an array of interoperability challenges (see sidebar).
Standardization has been achieved in high-speed data applications with the introduction of CableLabs DOCSIS certification testing for cable modems. However, DOCSIS certification procedures were not intended to test published performance limits or all specified functionality that may reside within different cable modems.
Before delving more into the standardization issue, it’s important to know a bit about CableLabs and DOCSIS. DOCSIS was created by Cable Television Laboratories (CableLabs), a research and development consortium of cable television systems operators representing North America and South America. CableLabs Certified® or CableLabs Qualified means that a device has passed a series of tests for compliance with the indicated version of the DOCSIS specification and demonstrated interoperable functionality with any other CableLabs certified/qualified device.
The certification testing process used by CableLabs to verify cable modem compliance to the DOCSIS 1.0 specification has evolved over the last several years into a very efficient, automated process. Three specifications make up DOCSIS 1.0: Radio Frequency Interface (RFI), Baseline Privacy Interface (BPI), and the Operations Support System Interface (OSSI).
The RFI specifications define the RFI for high-speed data applications over cable. The BPI specification describes a simple data privacy function for cable modem terminations systems (CMTSs) and is intended to provide a minimum level of data privacy and protection from theft for Internet services. The OSSI specification details the managed objects and management communications protocol and addresses the operations support requirements in a uniform and consistent manner.
The result of testing to these specifications is a high degree of interoperability among products from many vendors. CableLabs has extended the specifications to meet new requirements and ramped-up testing on the DOCSIS 1.1 version of the specification.
While this certification testing does an excellent job of ensuring interoperability of products from multiple vendors, it does not paint the complete picture. Areas such as product performance, robustness, and fault recovery are left to the vendor and individual operator. Additional testing in real-life HFC plant conditions is necessary to ensure adequate network performance.
ARRIS has conducted this additional testing for several years to augment the certification test process. A cable modem interoperability test (CMIT) lab was established to test the performance of cable modems from multiple vendors, using CMTSs headend products. Testing procedures typically focused on three areas:
- CMTS/cable modem basic operation and feature set compatibility.
- CMTS/cable modem performance.
- CMTS/cable modem RF characteristics.
Test procedures are run in both an optimum network configuration and one that simulates a real-life broadband HFC network. The system-under-test includes the CMTS and cable modems interfaced via the HFC network.
Industry-standard network traffic generators and protocol analyzers initiate and monitor Ethernet data traffic in a full range of traffic patterns. The patterns represent a diverse suite of operational service and application conditions offered throughout a cable-operator’s community.
On the RF side, an arbitrary waveform generator combined with a Gaussian noise generator provides source impairments that simulate real-world HFC environments. Operators can duplicate these tests as appropriate for the CMTS unit of their choice. To obtain test results that will reliably predict the operation of a specific cable modem in a network, test samples of 50 to 75 units are suggested.
Basic Operation and Feature Set Compatibility
A wide range of capabilities is exercised in this test area. Following is a high-level summary of the test areas and success criteria.
Test 1. Upstream Ingress Noise Robustness
This test suite verifies the capability of the cable modem to perform upstream ingress avoidance. This feature is enabled by a CMTS that detects upstream noise ingress and commands the cable modem to move to an alternative frequency, bandwidth, or transmission profile.
Test 2. Cable Modem Performance and Support of Multiple Upstreams
This test suite verifies the capability of the cable modem to range and register on the correct upstream with multiple upstream channel descriptors (UCDs) being sent on the downstream. The capability of the cable modem to be moved from one physical upstream to another, simulating an operator moving a cable modem from one fiber node to another, also is verified.
Data packets of various sizes (64 to 1,518 bytes) are generated on the Ethernet side of the CMTS and terminated on the Ethernet side of the cable modem. The breaking point for each cable modem across the range of packet sizes is determined by increasing the data rates. Each test is repeated with the direction of traffic reversed. Finally, all tests are repeated for all possible RF modulation profiles.
Test 3. Cable Modem Recovery From Power Outage
This test verifies the cable modem’s capability and speed in recovering from a network-wide power outage. It is performed with all eight upstream receiver cards in the CMTS active.
Test 4. Cable Modem Filter Performance
One key function performed by the cable modem is filtering various types of traffic before it enters the operator’s network. The cable modem also can perform filtering in the downstream direction.
Extensive Internet Protocol (IP) layer and Logical Link Control (LLC) filters testing is performed on all ports of the cable modem to verify adequate performance (Table 1). Figure 1 details an example lab setup for exercising cable modem filters.
| Test | Filtering Type | Criterion | Filter Port |
| F1 | Cable Modem-IP Address | Destination Address | Ethernet |
| F2 | Cable Modem-IP Address | Destination Address | RF |
| F3 | IP Address Inbound | Source Address | RF |
| F4 | IP Address Inbound | Source Address | Ethernet |
| F5 | IP Address Outbound | Source Address | RF |
| F6 | IP Address Outbound | Source Address | Ethernet |
| F7 | IP Port | Port Number | RF |
| F8 | IP | Transport Protocol | Ethernet |
| F9 | Cable Modem-LLC | LLC Field | RF |
| F10 | Cable Modem-LLC | DSAP Field | RF |
RF Characteristics
The RF characteristics test area covers items such as modem registration times, reboot times, impairment sensitivity, frequency allocation, and transmit/receive thresholds. These tests, shown in Table 2, are automated to accelerate the test cycle and increase test coverage across all possible frequency/modulation/power combinations. The capability to verify cable modem performance across a large spectrum of operating conditions greatly decreases the odds of finding costly compatibility issues during network deployment.
| Test | Name | Criteria |
| RF1 | Downstream | Signal to Noise—sensitivity to downstream broadband Gaussian noise at 256QAM; verifies accuracy of DocsIfSignalQuality MIB |
| RF2 | Rx Frequency Selectivity | Downstream Scan Time—frequency selectivity from 91 to 857 MHz at 64QAM and 256QAM |
| RF3 | Rx Level Sensitivity | Connectivity Loss—operation within DOCSIS receiver power window and additional robustness outside of this range |
| RF4 | Level Sensitivity | Synch Loss—immunity to abrupt changes in upstream/downstream plant levels |
| RF5 | Tx Frequency Selectivity | Modem Response Time—verifies that cable modem can transmit on a wide range of upstream frequencies and modulation profiles |
| RF6 | Tx Output Level Sensitivity | Connectivity Loss—measures across the entire DOCSIS transmit power dynamic range |
| RF7 | Impairment Test | Signal to Noise—sensitivity to downstream broadband Gaussian noise at 64QAM; verifies accuracy of DocsIfSignalQuality MIB |
Summary
The early success of cost-effective, high-speed data deployments over HFC can be directly attributed to the effectiveness of the CableLabs DOCSIS certification program. However, with the increased number of cable modem vendors, the growing size of network deployments, and increased reliance on this always-on service, additional system-level integration testing is required.
About the Author
Bruce McClelland joined ARRIS in 1999 and today is the vice president of engineering for ARRIS Broadband. Prior to joining ARRIS, he spent four years as director of engineering at Nortel in North Carolina and seven years with Nortel in Ottawa, Canada, as systems architect for the Broadband STP. Mr. McClelland received a B.E. from the University of Saskatchewan. ARRIS Broadband, 3871 Lakefield Dr., Suite 300, Suwanee, GA 30024, 770-622-8474, e-mail: [email protected]
FOR MORE INFORMATION
on cable modem/DOCSIS
www.rsleads.com/304ee-177
Return to EE Home Page
Published by EE-Evaluation Engineering
All contents © 2003 Nelson Publishing Inc.
No reprint, distribution, or reuse in any medium is permitted
without the express written consent of the publisher.
March 2003