HUAWEI UC Technologies and Standards Sales Specialist Training Topics Overview Basic Concepts of Enterprise Voice Systems Voice System Networks and Reliability VoIP Concepts and Protocols Basic Concepts of Video Basic Concepts of VoIP QoS VoIP and IP Networks 2 What Can IP Telephony & Unified Communications Do? Facilitate voice communication among users within enterprises Unified numbering within the dedicated network, simplifying memorization, and dialing More telephony services; for example, Calling Line Identification Presentation (CLIP) and callback ... Facilitate communication with customers Unique PBX number that is easy to remember The PBX can help customers reach a desired enterprise user ... Improve team collaboration capabilities Organization-matching functions, including searching a group, secretarial service, and call pickup. Voice and data conferencing functions ... 3 Save costs for enterprises VoIP calling saves toll call fees Save training fees Enable efficient call routing Billing ... Enhance work efficiency Integrate communication services into service systems ... Support enterprise business development Hotel Production and dispatching ... IP Telephony (IPT) and Unified Communications (UC) enable enterprises to improve work efficiency, service quality, and their enterprise image while reducing costs. Evolution of Enterprise Communications The first modernized Private Branch Exchange (PBX) or Private Automatic Branch Exchange (PABX) was developed in the middle 1970s, during the same period as Carriers’ “stored program control” exchanges. The PBX has changed in the following stages: IP-enabled PBX, IP PBX, and SIP-based IP PBX. Unified Communications (UC) was derived from PBX. Unified Communications and Collaboration (UC&C) evolved from UC. Commercially available Mature Commercially available Mature Commercially available Mature Commercially available Mature 1970s 1980s 1998 2008 2005 2015 2010 2020 PBX •More efficient communication tool •Reduced communication costs IPT •Beginning of IP era; reduced network costs •Further reduced communications costs •More flexible deployment 4 UC •Integrates communication tools •Changes communication modes •Improves working efficiency UCC/CEBP •Integrates communication services into service systems •Changes work modes •Improves Enterprise competitiveness Differences between Enterprise and Carrier Voice Communications Services PBXs concentrate on users’ experience, user information security, convenience of creating their own private networks, and ensure service quality over transmission networks with limited performance. PBXs have less impact on O&M and billing. Currently, there are hundreds of PBX services provided by mainstream vendors. However, Carriers can provide only some supplementary services. In addition to providing telephony services, PBXs: Permit enterprise users to make internal calls free of charge Help enterprises create private voice networks, reducing toll call fees Permit enterprises to control their own private voice networks Conveniently add applications such as voice mailbox, unified messaging, voice dispatching, emergency paging system, and hotel voice mail applications Easily interconnect with system functions such as broadcasting, electric control switch, alarms, and wireless trunking ... These advantages of PBXs ensure that enterprises’ voice communication services will be a growing market for a long time to come. 5 Benefits of Enterprise Communications Endpoints •Enterprise endpoints offer comprehensive functions. The popular IP phone provides some unique functions. However, most of the functions have been available since the digital phone era. •Using dedicated enterprise endpoints, users can manage audio services on PBXs simply by pressing buttons. This frees users from memorizing complicated function access codes. The price of a medium-range endpoint ranges from USD $100 to USD $200. 6 Topics Overview Basic Concepts of Enterprise Voice Systems Voice System Networks and Reliability VoIP Concepts and Protocols Basic Concepts of VoIP QoS VoIP and IP Networks 7 Telephone Switching Addressing and connecting the call Placing a call Answering the call Program control era: converting from analog to digital and from manual to computerized IP era: converting from Time Division Multiplex (TDM) to Packets (basic principles and structure have not changed). 8 PSTN and PBX Public Switched Telephone Network (PSTN) is a telecommunications network established to perform telephone services for public subscribers PBX A device on the user side from a Carrier’s perspective Evolved from the Integrated Services Digital Network (ISDN) in the TDM era, and quickly transitioned from H.323 to Session Initiated Protocol (SIP) in the IP era Analog POTS Handset Trunks POTS Handset Trunks Trunks SS7 PSTN PBX 9 SS7 Switch B ISDN PRI/BRI Switch A POTS Handset Switch C Signaling System No. 7 (SS7) •Communications devices follow a specified communications protocol to transmit control information to destination devices in a secure, reliable, and efficient manner. The transmitted information is called “protocol control information” in computer networks, and “signal” or “signaling” in telecommunications networks. •Signaling is divided into subscriber signaling and inter-office signaling. Subscriber signaling is used between user terminals (for example, telephones and PBXs) and the PSTN, while inter-office signaling is used within the PSTN. •SS7 is an inter-office signaling for Carriers and is termed Common Channel Signaling (CCS). SS7 plays a crucial role in intelligent communication network development. ITU-T definition • SS7 core modules Telephone User Part (TUP) ISDN User Part (ISUP) Only Huawei’s PBXs support SS7, which is our bidding specification advantage in China. 10 R2 Signaling and CNo.1 Signaling R2 signaling is channel-associated signaling. It uses a signaling channel associated with the speech channel or the speech channel itself to transmit the required control signals such as a busy signal, answer signal, release signal, and dialing signal. In other words, R2 signaling uses one channel to transmit both speech information and associated signaling. China No.1 (CNo.1) signaling is a subset of the International R2 signaling system and is widely used by PSTN networks in China Outside of China, Mexico, and Brazil, R2 signaling is a special subset of the International R2 signaling system Inter-office signaling and user-side signaling are not differentiated with R2, so all mainstream PBXs support R2 R2 signaling has simple functions, poor scalability, low efficiency, and small capacity, so R2 signaling has been totally replaced by SS7 for PSTN 11 E1 and T1 E1 and T1 are standards for transmitting data over physical lines and are used for signaling. E1 is dominant in Europe. T1 is dominant in the U.S., Canada, Hong Kong, Taiwan, and Japan (named J1 by some vendors). Similarities: the same sampling frequency (8 kHz), bits per code (8 bits), and timeslot bit rate (64 kbit/s) Differences: an E1 has 32 timeslots and a data transmission rate of 2.048 Mbit/s. A T1 has 24 timeslots and a data transmission rate of 1.544 Mbit/s. E1 adopts A-law coding/decoding of 13-segment while T1 adopts µ-law coding/decoding of 15-segment. Interfaces: unbalanced 75 ohm coaxial cable and balanced 120 ohm twisted pair based on G.703. 12 PRI/BRI interface PRI and BRI are used for subscriber signaling and are also called Digital Signal System 1 (DSS1) signaling. PRI and BRI are products of the ISDN era and final products of the TDM voice era. The PSTN is used to connect PBXs or endpoints Primary Rate Interface/Primary Rate Access (PRI/PRA): 30B+D interface Basic Rate Interface/Basic Rate Access (BRI/BRA): 2B+D interface and two types of physical interfaces S/T interface: 4 lines and a transmission distance of 1.2 km (the BRI interfaces in some countries are S/T interfaces) U interface: 2 lines, a transmission distance of 5 km, and converted to S/T interface through NT1 Type B Channel (Media, kbit/s) D Channel (Signaling, kbit/s) BRA 2B+D 2B = 2 x 64 = 128 D = 16 PRA Europe 30B+D 30B+D = 30 x 64 = 1,920 D = 64 PRA North America 23B+D 23B+D = 23 x 64 = 1,472 D = 64 13 Q Signaling (QSIG) QSIG is a protocol for Integrated Services Digital Network (ISDN) communications based on the Q.931 standard, and is used for signaling between PBXs QSIG supports a variety of functions such as basic calling, number display, name display, call transfer, call forwarding, call back, message notification, and route optimization 14 FXO and FXS FXO FXS PSTN PBX Foreign Exchange Office (FXO), also called an analog trunk, is an interface on a PBX to connect to the PSTN. The FXO interface helps simulate the PBX as an analog phone to interact with Carrier networks. In FXO connection mode, users can connect to extensions only through an operator service or the automatic switchboard. Foreign Exchange Station (FXS), also called a Plain Old Telephone Service (POTS) port, is an interface on the PBX to connect to analog phones. This interface supplies battery power, provides dial tone, and generates ringing voltage. 15 Differences between Digital Phones and IP Phones Digital Phone IP Phone Physical connection Twisted pair Ethernet cable Power supply Centralized PoE or local power supply Transmission protocol 2B+D IP IPT service Comprehensive Comprehensive+ Bandwidth service N/A Imaginative Transmission distance A maximum of 1,200 meters Distance that IP networks can reach Power-off survival Unsupported Unsupported Phone relocation Troublesome Simple Extension mobility Partly supported Supported Compatibility Self-developed and incompatible High compatibility with basic services Hardware specification Low hardware configurations (for example, screen and Bluetooth are not supported) Technology keeps current with industry 16 Digital phones and IP phones do not have significant differences in appearance, but may have three versions: digital phone, H.323, or SIP. Topics Overview Basic Concepts of Enterprise Voice Systems Voice System Networks and Reliability VoIP Concepts and Protocols Basic Concepts of Video Basic Concepts of VoIP QoS VoIP and IP Networks 17 Traditional TDM PBX (Digital SPC) E1/T1/BRI/SS7/R2/ PRI digital trunk User cable PSTN Equipment room MDF FXO analog trunk Cables are the main disadvantage for a traditional TDM because of investment costs and maintenance. Security and reliability of cables are also poor. For example, cables are prone to lightning strikes. The advantages of analog phones are that power can be supplied in a unified manner, and phones can be bridged. SPC Floor and building distribution frame Twisted pair phone cable 18 Multiple twisted pairs phone cable Analog User Access with IP PBX E1/T1/BRI/SS7/R2/ PRI digital trunk IP PBX User cable PSTN Equipment room MDF FXO analog trunk Multiple twisted pairs phone cable Using Integrated Access Devices (IADs) to provide IAD access for analog users, IP PBX eliminates the Twisted pair phone cable disadvantage of user cables. 19 Floor and building distribution frame IP User Access with IP PBX E1/T1/BRI/SS7/R2/ PRI digital trunk IP PBX PSTN Switch FXO analog trunk IP network IP phones can be powered by switches that provide Floor and building switch Power over Ethernet (PoE), or use local power supplies IP phones cannot be bridged 20 Integrated cabling Topics Overview Basic Concepts of Enterprise Voice Systems Voice System Networks and Reliability VoIP Concepts and Protocols Basic Concepts of Video Basic Concepts of VoIP QoS VoIP and IP Networks 21 Main VoIP Standards: H.323 and SIP H.323 is a general standard developed by the International Telecommunication Union (ITU) for sharing audio, video, and data over data packet (IP) networks. Initially, H.323 was used in multimedia conferences. Later, it was expanded and used by IP phones. Has typical telecom features with advantages from TDM development to IP for Carrier networks Provides unified processing and management SIP is a multimedia signal protocol developed by the Internet Engineering Task Force (IETF). Simple, modular, good expandability, and closely associated with Internet applications Pushes the complexity of network devices to the edge of the network Focuses on building VoIP networks based on the Internet Common ground between H.323 and SIP: Provide multimedia services over IP networks Run on IP networks, use TCP, and UDP sessions to transmit signals; use RTP to transmit voice and video streams Use existing protocols (such as G.711 and G.729) for encoding and decoding; do not require new encoding/decoding methods Use a server as an intermediary for establishing sessions • H.323 VoIP network: a gatekeeper provides address translation, bandwidth control, certificate control, and area management functions • SIP network: an agent server processes and sends requests of user agents, directly establishes sessions with other user agents, and calls traditional PSTN users through a gateway 22 Differences between H.323 and SIP An important feature of SIP is that it does not define the types of sessions to be established. SIP defines only the method for managing sessions. Based on this flexibility, SIP can be used in diversified applications and services, including interactive games, on-demand music and video applications, and for voice, video, and data conferences. By design, SIP is a distributed call model, and provides a distributed multicast function. The multicast function not only facilitates conference control, but also simplifies user positioning, group invitation, and saves bandwidth. H.323 does not support multicast. Advantages of SIP: Simple, easy to understand: SIP messages are in text format, while H.323 messages are in ASN.1 format Extensible: many parts of SIP can be customized by users, while H.323 cannot Expandable: SIP supports multi-domain searches High efficiency: the process for establishing a SIP call is easier than for H.323 The cost for establishing an audio and video environment is low Disadvantage of SIP: The simpler, the more popular Lack of standards 23 Digitization Basics: PCM Voice Encoding Pulse Code Modulation (PCM) samples analog signals such as voice or image signals periodically to make them discrete, round, and quantized sample values. Then, sample values are converted to binary code to represent amplitude values of sample pulses. Encoding process Sampling Quantizing Coding Traditionally, voice encoding uses an 8 kHz sampling rate, an 8-bit depth to coding quantized values, and uses A-law or µ-law in the coding process to finally obtain 64 kbit/s voice code. 24 Opus Opus is a lossy audio codec, which was developed by the IETF for real-time voice transmission over networks. In the competition among lossy audio formats, Advanced Audio Coding (AAC) was once very popular. However, Opus quickly upstaged AAC. In low-bit-rate encoding, Opus outperforms HE AAC. In middle-bit-rate encoding, Opus is comparable to AAC with 30% higher bit rate. In high-bit-rate encoding, Opus is closer to the original voice. Therefore, Opus has broad uses in the future. 6 kbit/s to 510 kbit/s bit rate 8 kHz (narrowband) to 48 kHz (full-band) sampling rate 2.5 ms to 60 ms frame size Constant Bit Rate (CBR) and Variable Bit Rate (VBR) support Audio bandwidth from narrowband to full-band Voice and music support Monaural and stereo support Up to 255 channels (frames of multiple data streams) Dynamically adjustable bit rate, audio bandwidth, and frame size Good robustness and Packet Loss Compensation (PLC) Floating point and fixed-point implementation When NetATE is enabled, packetization time and bit rate are automatically adjusted based on network conditions. This applies only to Opus. Therefore, the bandwidth required by Opus is not fixed. 25 Topics Overview Basic Concepts of Enterprise Voice Systems Voice System Networks and Reliability VoIP Concepts and Protocols Basic Concepts of Video Basic Concepts of VoIP QoS VoIP and IP Networks 26 International Standards for Video Encoding Moving Picture Experts Group (MPEG) standards: include MPEG-1, MPEG-2, and MPEG-4 standards developed by the International Organization for Standardization (ISO) and International Electrotechnical Commission (IEC) ITU-T standards: include H.261 and H.263 developed by ITU-T for video phones and conferences H.264/Advanced Video Coding (AVC) (MPEG-4 Part 10) jointly developed by the ISO and ITU-T 27 H.263 (Past) H.263 is a low-bit-rate video codec developed by the ITU-T for video conferences. Initially, H.263 was designed to transmit data based on H.324 systems (that is, conducting video conferences or calls over a PSTN or other networks that are based on circuit switching). Later, it was found that H.263 also can be successfully applied in H.323 (video conference system based on an RTP/IP network), H.320 (video conference system based on an ISDN), RTSP (streaming media transmission system), and SIP (video conference system based on the Internet). H.263 provides better image quality than H.261 (designed for ISDN) in low bit rates. Here are the differences between H.263 and H.261: H.263 motion compensation uses half-pixel precision, while H.261 uses full-pixel precision and loop filter. Some parts of the data stream hierarchy structure are optional in H.263. This allows a lower bit rate or better error correction to be configured for H.263. H.263 contains four negotiable options to improve performance. H.263 uses unrestricted motion vectors and syntax-based arithmetic encoding. H.263 uses the same frame prediction method as the P-B frame in MPEG. H.263 supports five resolutions: in addition to Quarter Common Intermediate Format (QCIF) and Common Intermediate Format (CIF) supported by H.261, H.263 also supports Sub-Quarter Common Intermediate Format (SQCIF), 4 x Common Intermediate Format (4CIF), and 16 x Common Intermediate Format (16CIF). The resolution of SQCIF is half that of QCIF, while the resolutions of 4CIF and 16CIF are 4 and 16 times that of CIF, respectively. 28 H.264 (Now) After H.263, the next-generation video codec developed by the ITU-T (jointly with the MPEG) was H.264, which is also called AVC or MPEG-4 Part 10. H.264 introduces many new compression technologies such as multiple reference frames, multi-block type, integral transform, and intra-frame prediction, and uses finer sub-pixel motion vectors (1/4 and 1/8) and a next-generation loop filter to improve compression performance and provide a complete system. Compared with H.263+ and MPEG-4 SP, H.264 saves up to 50% of the bit rate, greatly reducing storage capacity. H.264 provides better video quality in various resolutions and bit rates. H.264 adopts a simple and clear design, uses simple syntax description, avoids excess options and configurations, and utilizes existing encoding modules (as many as possible). H.264 can be easily combined with low-bit-rate codecs such as G.729 for a complete system. H.264 features low delay and flexibly uses appropriate delay limits for different services. H.264 enhances error code and packet loss processing to improve the decoder’s error correction. H.264 provides higher network adaptability, using network infrastructure and syntax to adapt to both IP and mobile networks as well as applications. 29 H.265 High-Efficiency Video Coding (Future) Huawei holds various core patents built on H.265, and is the dominant player. H.265 aims to transmit network videos with higher quality using limited bandwidth. Using half the bandwidth originally required, H.265 can provide videos with the same quality. Compared with H.264, H.265 can reduce video size from 39% to 44%, while ensuring the same video quality. When the bit rate is reduced by 51% to 74%, H.265 can still provide video quality equal to or even better than H.264. Essentially, H.265 provides better Peak Signal to Noise Ratio (PSNR) than expected. H.265 also supports ultra HD videos such as 4k (4096 x 2160) and 8k (8192 x 4320) videos. H.263 can transmit SD broadcast digital TV videos (720 x 576 that comply with CCIR601 and CCIR656) at a bandwidth of 2 Mbit/s to 4 Mbit/s. Based on algorithm optimization, H.264 can transmit SD digital images at bandwidth lower than 2 Mbit/s. H.264 HD can transmit 1080p full HD videos at a bandwidth lower than 1.5 Mbit/s. 30 Topics Overview Basic Concepts of Enterprise Voice Systems Voice System Networks and Reliability VoIP Concepts and Protocols Basic Concepts of Video Basic Concepts of VoIP QoS VoIP and IP Networks 31 Voice Quality Factors Using VoIP Common methods for improving voice quality: • Packet Loss Compensation (PLC) • Dynamic Jitter Buffer (DJB) • Automatic Echo Cancellation (AEC) • Automatic Noise Suppression (ANS) • Automatic Silence Compression (ASC) • Voice Activity Detection (VAD) • Comfort Noise Generation (CNG) • Automatic Gain Control (AGC) Voice quality evaluation method: Mean Opinion Score (MOS) test Main factors affecting VoIP voice quality Now we can see why vendors do not guarantee VoIP voice quality over 3G, 4G, Wi-Fi networks, or the Internet! 32 Level MOS Score User Experience Excellent 5.0 Very clear, no distortion, and no delay Good 4.0 Clear, low delay, and a little noise Fair 3.0 Not very clear, and obvious delay, noise, and distortion Poor 2.0 Not very clear, loud noise, intermittent, and serious distortion Bad 1.0 Silent or absolutely unclear, and very loud noise User Experience with Delay Delay defined by the ITU in G.114 Range (ms) Description 0–150 Acceptable to most users and applications 150–400 Acceptable in certain conditions > 400 Unacceptable User experience Range (ms) Description 0–100 Not felt by most people 100–300 Slightly affected > 300 Obvious feel Note: Delay refers to end-to-end delay. Considering other factors, such as encoding/decoding and buffering, network delay should be controlled within 200 ms. 33 Packet Loss Concealment (PLC) Algorithm Missing frame Restored frame Packet loss detector 1 or 0 Frame estimator Frame buffer 4 5 6 8 9 Frame count Feature vector stream from network To recognizer VoIP transmits voice data over an IP network using UDP. On an IP network, packet loss will unavoidably occur. To minimize effects from packet loss, the PLC algorithm can be used to reconstruct missed frames based on the correlation inside voice information, ensuring received voice quality. If the packet loss rate is high, not all missed frames can be compensated through calculation. However, voice services are not as sensitive to packet loss. 34 Dynamic Jitter Buffer (DJB) PBX MG IP network Jitter buffer 20 ms 20 ms 20 ms 23 ms 20 ms 20 ms C B A C B A C B A 50 30 10 50 30 10 50 30 10 RTP timestamp 20 ms frame interval IP users RTP timestamp 20 ms frame interval RTP timestamp 20 ms frame interval A dynamic jitter buffer refers to a certain size buffer allocated by a gateway’s RTP media receiver, where RTP media packets are buffered, sorted, and discarded. The main function of the buffer is to reduce the impact of network jitter on voice service. When there is no network jitter, the buffer can be disabled; when network jitter is large, the size of the buffer can be increased. Buffering results in network delay. A larger delay is better to filter jitter. A DJB using an excellent algorithm can find the balance between delay and jitter, and enable good queuing and timely discards to obtain perfect Internet voice quality. 35 Topics Overview Basic Concepts of Enterprise Voice Systems Voice System Networks and Reliability VoIP Concepts and Protocols Basic Concepts of Video Basic Concepts of VoIP QoS VoIP and IP Networks 36 Power Supply Methods for IP Phones Independent power module Customer LAN Desktop PC switch This situation illustrates what to do when a customer’s LAN switch is old. The wall-mounted power supply for IP phones must have Uninterruptible Power Supply (UPS) capability. A PoE module can be directly attached to a LAN switch port for use with old devices. Desktop PC Any PoE switch that complies with the 802.3af PoE standard or media gateway with a built-in PoE switching port Mainstream IP phones have PoE capability. PoE enables UPS deployment and LAN switch requirements in a unified manner, which simplifies desktop cabling. 37 Why Don’t H.323 and SIP Packets Support NAT Traversal? H.323 and SIP protocol packets write original address information at the application layer. Network Address Translation (NAT) only converts addresses at the network layer. When the destination end receives a packet, and finds that the address at the application layer (original address) is different than the address at the network layer (address after NAT), the destination end discards the packet. A Session Border Controller (SBC) functions as the NAT device at the application layer. An SBC is required because it converts addresses at the network layer in addition to the application layer. 38 HUAWEI ENTERPRISE ICT SOLUTIONS A BETTER WAY Copyright © 2014 Huawei Technologies Co., Ltd. All Rights Reserved. The information in this document may contain predictive statements including, without limitation, statements regarding the future financial and operating results, future product portfolio, new technology, etc. There are a number of factors that could cause actual results and developments to differ materially from those expressed or implied in the predictive statements. Therefore, such information is provided for reference purpose only and constitutes neither an offer nor an acceptance. Huawei may change the information at any time without notice.