Fact-checked by Grok 2 weeks ago

Circuit switching

Circuit switching is a method of telecommunications networking in which a dedicated end-to-end communications path, or circuit, is established between two nodes before data transmission begins and remains reserved exclusively for their use throughout the entire duration of the session.^[1] This approach guarantees a fixed amount of bandwidth and minimal latency variation, making it ideal for applications requiring consistent, real-time performance, such as traditional voice telephony.^[2] The origins of circuit switching trace back to the early development of telephone systems in the late 19th century. In January 1878, the first telephone switch was deployed in New Haven, Connecticut, allowing multiple users to share physical lines by establishing temporary dedicated circuits on demand.^[3] A major advancement came in 1891 with Almon Strowger's invention of the automatic electromechanical switch, which automated the connection process and reduced reliance on human operators, paving the way for scalable public switched telephone networks (PSTN).^[4] Over the 20th century, the technology evolved from manual and analog electromechanical systems to digital implementations using techniques like time-division multiplexing (TDM) and pulse-code modulation (PCM), particularly accelerating in the 1960s and 1970s as global telephone infrastructure expanded.^[4] In operation, circuit switching proceeds through three distinct phases: setup, where signaling protocols (such as those defined by early standards from the International Telecommunication Union) identify the destination and allocate a continuous path across switches and transmission links; conversation or data transfer, during which the full circuit capacity is devoted to the communicating parties with no sharing or contention; and teardown, which releases the resources once the session ends.^[1] Key characteristics include connection-oriented service, traffic isolation to prevent interference, and support for multiplexing methods like frequency-division (FDM) or time-division (TDM) to combine multiple circuits over shared media.^[1] While it offers advantages such as predictable quality of service (QoS), high reliability with rapid fault recovery (e.g., under 50 milliseconds in systems like SONET/SDH), and simplified processing at network nodes, circuit switching is criticized for bandwidth inefficiency—resources remain idle during silences or pauses—and inflexibility in handling bursty or variable traffic patterns.^[1] As of 2025, it continues to be used in some legacy PSTN cores and certain high-capacity backbone links, but is being phased out in many regions, including the completion of the PSTN switch-off in the UK by December 2025, in favor of packet switching in data-centric networks like the internet.^[2]^[5]

Fundamentals

Definition and Principles

Circuit switching is a method of implementing telecommunications networks in which a dedicated communications channel, or circuit, is established between two network nodes prior to data transmission, reserving the path exclusively for the duration of the session to guarantee constant bandwidth.^[6] This technique originated in early telephony systems and involves pre-allocating network resources, such as bandwidth and switching capacity, from the source to the destination, ensuring uninterrupted connectivity once the circuit is active.^[7] The basic principles of circuit switching revolve around establishing end-to-end connectivity via intermediate switches that configure the dedicated path, preventing resource sharing during the active communication phase.^[8] Unlike shared-medium approaches, which multiplex multiple signals over common resources, circuit switching allocates the full capacity of the path to a single connection, minimizing interference but potentially leading to underutilization if the circuit is idle.^[9] This dedicated allocation supports reliable transmission without the need for dynamic routing decisions during data flow. Key characteristics of circuit switching include fixed delay, as the path is predetermined and resources are reserved in advance, and constant bit rate, which maintains a steady data flow rate throughout the session.^[10] These properties make it particularly suitable for real-time applications like voice and video telephony, where low jitter and predictable performance are essential, as opposed to bursty or variable-rate data traffic.^[11] Circuits in this paradigm can be realized through space-division methods, which employ physical wiring or crosspoint matrices to create separate paths for each connection, or time-division multiplexing (TDM), which implements virtual circuits by assigning fixed time slots within a shared high-speed frame to simulate dedicated channels.^[9] Space-division approaches provide true physical isolation, while TDM enables efficient use of digital transmission lines by interleaving samples from multiple circuits without spatial separation.^[8]

Circuit Components and Types

Circuit-switched networks rely on several core components to establish and maintain dedicated paths for communication. Switches form the backbone, enabling the routing of connections between endpoints; common types include crossbar switches, which use a matrix of crosspoints to interconnect multiple lines electronically or electromechanically, rotary switches (also known as step-by-step or Strowger switches), which mechanically select paths by rotating wipers to specific contacts, and electronic switches, which employ digital logic for faster, non-mechanical crosspoint selection in modern systems. Transmission media provide the physical pathways, primarily copper wires such as twisted-pair cables for local loops in traditional telephony, and fiber optics for high-capacity, long-distance trunks that support greater bandwidth and reduced signal attenuation. Signaling systems handle control functions like call setup and supervision; a prominent example is Signaling System No. 7 (SS7), a protocol suite that enables out-of-band communication between network elements to manage circuit allocation and routing in public switched telephone networks. Circuits in these networks vary by allocation and signal nature. Switched circuits are dynamically established on demand for the duration of a session, allocating resources only when needed and releasing them afterward, whereas permanent circuits are pre-provisioned dedicated paths, such as leased lines, that remain active indefinitely for continuous use without per-call setup. Analog circuits transmit continuous varying signals, suitable for early voice telephony where waveforms are directly modulated onto carriers, while digital circuits convert signals to discrete binary samples, enabling multiplexing and error correction but requiring analog-to-digital conversion at endpoints. Multiplexing techniques optimize resource use by combining multiple circuits onto shared media. In analog systems, frequency-division multiplexing (FDM) divides the available bandwidth into non-overlapping frequency bands, each assigned to a separate channel with guard bands to prevent interference, allowing simultaneous transmission of voice signals over coaxial or microwave links. For digital systems, time-division multiplexing (TDM) interleaves signals by allocating fixed time slots within repeating frames; a basic TDM frame structure consists of a sequence of slots, where each slot carries a sample from one channel (e.g., 8-bit pulse-code modulation samples for voice), synchronized across the network to ensure orderly reconstruction at the receiver. Network topology in circuit-switched systems influences routing efficiency and scalability, often employing hierarchical star configurations where local switches connect end users to central offices, which in turn link to tandem switches for inter-office routing, minimizing cabling while centralizing control. Mesh configurations may interconnect higher-level switches for redundancy and direct paths between major nodes, though they increase complexity compared to pure star hierarchies.

Operation

Call Establishment

Call establishment in circuit switching begins when the originator initiates a connection by dialing the destination address, prompting the local switch to propagate a setup signal through intermediate switches to reserve a dedicated path of circuits. This signaling message travels hop-by-hop, with each switch allocating a channel and updating its connection table if resources are available, until it reaches the destination switch, which confirms the end-to-end path via backward signaling to notify the originator and complete the reservation.^[12]^[13] Signaling protocols during call establishment can be in-band, utilizing the same voice channel for control messages, or out-of-band, employing a separate channel to avoid interference with data transmission. In-band examples include DTMF tones generated by the caller's device to convey digits to the local switch.^[12] Out-of-band signaling, such as common channel signaling via SS7, uses a dedicated packet-switched network for efficient, high-speed control between switches, enabling global coordination without occupying the bearer channels.^[12] Path selection algorithms determine the route during setup, with fixed routing predefining paths based on network topology and static link weights for simplicity and predictability. Least-cost routing, in contrast, dynamically selects the path minimizing a cost metric, such as propagation delay or congestion, by checking criteria like available bandwidth at each switch before reservation.^[13]^[12] If no end-to-end path with sufficient resources is found due to congestion, the call faces blocking probability, where the request is rejected to maintain quality for active connections. This probability is modeled by the Erlang B formula for traffic engineering in systems with m trunks and offered load A in Erlangs:

B(A, m) = \frac{A^m / m!}{\sum_{k=0}^{m} A^k / k!}

This quantifies the fraction of calls blocked in a loss system without queuing, guiding network dimensioning to keep blocking below targets like 1-2%.^[14]^[12]

Transmission Phase

Once the circuit is established, the transmission phase involves continuous and uninterrupted data flow over the dedicated end-to-end path, where the entire bandwidth is reserved exclusively for the communicating parties. This setup ensures a fixed data transmission rate, known as constant bitrate (CBR), which is particularly suited for real-time applications requiring predictable performance. In the Public Switched Telephone Network (PSTN), voice channels operate at a standard bitrate of 64 kbps, derived from pulse-code modulation (PCM) encoding at 8 bits per sample and 8,000 samples per second, enabling synchronous transfer without interruptions from other traffic.^[15]^[16] Error handling during transmission is primarily confined to the physical layer, focusing on analog-domain issues rather than digital packet-level mechanisms. For instance, echo suppression devices are employed to attenuate unwanted reflections in the circuit, such as talker echo caused by impedance mismatches in hybrid transformers, by inserting loss in the return path when far-end speech is absent. Unlike packet-switched networks, there is no retransmission of erroneous data at higher layers; instead, reliability depends on the inherent stability of the dedicated physical connection and basic signal conditioning.^[17] Synchronization between sender and receiver is maintained through clock recovery circuits that extract timing information from the incoming signal, ensuring precise alignment for data sampling. This process supports jitter-free delivery, critical for isochronous traffic like uncompressed voice, where variations in delay could degrade quality; the dedicated path guarantees constant latency without buffering or queuing delays. Bandwidth utilization remains fixed at the allocated rate throughout the session, but this leads to inefficiency if no data is transmitted, as the resources are held idle and unavailable to others, contributing to overall network underutilization during periods of silence.^[16]^[18]

Circuit Teardown

The circuit teardown phase in circuit switching begins when one endpoint signals the end of communication, typically through an action such as hanging up the telephone (on-hook signal), which initiates the disconnection process across the network.^[19] This release sequence propagates backward through the switches along the established path, ensuring that the dedicated resources are systematically freed for reuse by other connections. The originating switch detects the endpoint's signal and coordinates with intermediate and destination switches to dismantle the circuit, preventing resource wastage in the network.^[20] In modern telephone networks, teardown relies on signaling protocols like Signaling System No. 7 (SS7), where the originating switch sends a Release (REL) message to the destination switch, identifying the specific trunk circuit associated with the call.^[20] The destination switch responds with a Release Complete (RLC) message, confirming the disconnection and returning the circuit to an idle state.^[21] For abrupt disconnections, such as sudden line faults or endpoint failures, SS7 incorporates timers to detect non-responses and automatically initiate circuit release, ensuring network stability without manual intervention.^[22] Resource recovery during teardown involves de-allocating the reserved bandwidth, switch ports, and trunk lines, with network elements updating their state tables to mark the circuit as available.^[20] This process frees the entire end-to-end path, allowing immediate reuse for new calls and maintaining efficient spectrum utilization in bandwidth-constrained systems like the Public Switched Telephone Network (PSTN).^[23] Potential issues in circuit teardown include delays in call clearing due to signaling propagation or timer expirations, which can temporarily hold resources and reduce network throughput. In rare cases of incomplete release, brief glitches such as momentary crosstalk between lines may occur if switches fail to fully isolate the path before reallocation, though modern protocols minimize this through confirmation mechanisms like RLC.^[20]

Comparisons

With Packet Switching

Circuit switching and packet switching represent two fundamental paradigms for network resource management and data transmission. In circuit switching, a dedicated end-to-end path is established and reserved for the duration of the communication session, employing techniques such as frequency-division multiplexing (FDM) or time-division multiplexing (TDM) to allocate fixed bandwidth exclusively to that connection.^[24] This contrasts with packet switching, which employs statistical multiplexing to dynamically share network resources among multiple users; data is divided into packets that are routed independently via store-and-forward mechanisms, without reserving a fixed path.^[24] As a result, circuit switching ensures a consistent, predictable connection but can lead to underutilization during idle periods, while packet switching optimizes efficiency through on-demand resource allocation but introduces variability in path selection.^[24] Performance differences between the two arise primarily from their handling of latency and throughput. Circuit switching provides low and fixed latency once the circuit is established, consisting mainly of propagation delay across the fixed path plus transmission time, with no queuing delays; the end-to-end delay can be approximated as d = t_{\text{prop}} + t_{\text{trans}} + t_{\text{setup}}, where t_{\text{prop}} is the propagation time, t_{\text{trans}} = L / R (with L as packet length and R as link rate), and t_{\text{setup}} is a one-time overhead.^[24] In contrast, packet switching incurs variable latency due to per-hop processing, queuing, and dynamic routing, yielding an end-to-end delay of d = \sum (t_{\text{proc}} + t_{\text{queue}} + t_{\text{trans}} + t_{\text{prop}}) across N hops, which can exceed circuit switching delays under congestion but offers higher overall throughput for bursty traffic through efficient sharing.^[24] Circuit switching thus delivers variable throughput limited by the reserved bandwidth but avoids congestion, whereas packet switching achieves high efficiency for intermittent data flows at the cost of potential jitter and delays.^[24] These distinctions influence their suitability for specific use cases. Circuit switching excels in applications requiring constant-bit-rate (CBR) traffic, such as traditional voice telephony, where predictable low latency and guaranteed bandwidth prevent disruptions in real-time communication.^[24] Packet switching, however, is ideal for bursty, data-oriented traffic like internet browsing or file transfers, as it accommodates variable rates without wasting resources during silence periods, though it may suffer packet loss or reordering without additional protocols.^[24] While circuit switching's dedication avoids congestion but performs poorly for sporadic traffic—leading to idle resource waste—packet switching's flexibility enables better utilization for modern data networks but can degrade under heavy load without quality-of-service mechanisms.^[24] Hybrid approaches, such as circuit emulation services (CES), bridge these paradigms by encapsulating circuit-based traffic (e.g., TDM signals) into packets for transport over packet-switched networks like IP or ATM, preserving timing and CBR characteristics without full circuit reservation.^[25] This enables legacy voice or leased-line services to leverage packet infrastructure cost-effectively.^[26]

With Message Switching

Message switching is a communication technique in which an entire message is treated as a single unit, buffered at each intermediate node, and forwarded to the next node only after complete reception, utilizing a store-and-forward mechanism.^[27] In contrast, circuit switching establishes a dedicated end-to-end physical path for continuous data transmission, reserving bandwidth exclusively for the duration of the connection without intermediate storage.^[28] This fundamental difference results in circuit switching providing a steady stream of data with minimal variability, while message switching handles the message holistically but introduces delays at each hop due to buffering. Key contrasts between the two include the reservation strategy and handling of data flow: circuit switching requires upfront end-to-end resource allocation, ensuring predictable latency but preventing other users from utilizing the reserved path during idle periods, whereas message switching employs hop-by-hop storage and forwarding without dedicated paths, allowing dynamic sharing of links but subjecting messages to queuing delays based on node availability.^[27] Circuit switching avoids buffering-related delays, making it suitable for constant-bit-rate applications, but message switching better accommodates variable message sizes by not fragmenting data, though it demands more storage at nodes.^[28] Regarding efficiency, circuit switching can waste bandwidth during idle times on the dedicated path, leading to underutilization in bursty or intermittent traffic scenarios, while message switching incurs higher overhead from queuing and storage but enables more flexible resource sharing across multiple messages.^[29] Throughput in circuit switching achieves full bandwidth utilization when active, providing consistent performance for ongoing transmissions; in message switching, throughput depends on buffer availability and network congestion, often resulting in lower effective rates due to propagation and processing delays at each hop.^[29] Analysis shows message switching outperforms circuit switching for bursty traffic by avoiding fixed reservations, but circuit switching is superior for smooth, continuous flows.^[29] Historically, message switching originated in early telegraphy systems, where operators at intermediate stations stored incoming messages via Morse code until the next line was free, then forwarded them in a store-and-forward manner to manage limited bandwidth over long distances.^[30] This approach contrasted with the later adoption of circuit switching in telephony, which enabled direct, real-time voice connections without manual relaying, marking a shift toward dedicated paths for higher-speed, user-friendly communication.^[30] Message switching's store-and-forward paradigm later influenced the evolution of packet switching by introducing the concept of independent message handling.^[27]

History

Origins and Early Adoption

Circuit switching emerged as a fundamental technology in telecommunications following the invention of the telephone by Alexander Graham Bell, who received a U.S. patent for his device on March 7, 1876, creating an urgent need for scalable systems to connect multiple users without constant manual intervention.^[31] Early telephony relied on manual switchboards operated by human attendants, but rapid growth in subscriber numbers demanded more efficient, automated methods to establish dedicated circuits for voice communication.^[32] The origins of circuit switching are closely tied to the development of automatic telephone exchanges, pioneered by Almon Brown Strowger, an undertaker from Kansas City, Missouri. In 1888, Strowger conceived the step-by-step switch to automate call routing, motivated by his suspicion that manual operators were diverting business calls to competitors; he received a patent for this electromechanical device (U.S. Patent No. 447,918) on March 10, 1891.^[33] The Strowger switch used rotating and stepping mechanisms to connect callers via dedicated paths, marking the first practical implementation of circuit switching in telephony and replacing human operators for local calls.^[32] Early adoption began in the 1890s with the formation of the Strowger Automatic Telephone Exchange Company in 1891, which installed the world's first commercial automatic exchange in La Porte, Indiana, in 1892, serving 75 subscribers.^[34] By the mid-1890s, step-by-step switches were deployed in small U.S. communities, enabling direct dialing and reducing reliance on operators, though initial systems were limited to lines under 10,000 due to mechanical constraints.^[32] The technology spread internationally, with private automatic branch exchanges appearing in London, Ontario, Canada as early as 1893.^[35] Growth accelerated in the early 20th century through advancements in electromechanical switching, including the Bell System's contributions. The American Telephone and Telegraph Company (AT&T), part of the Bell System, initially favored manual methods but began integrating automatic switches after acquiring independent exchanges; its first step-by-step installation occurred in Norfolk, Virginia, in 1919.^[32] In the 1920s, Bell developed the panel switching system, an electromechanical improvement over Strowger's design for larger urban networks, with the first installation in Newark, New Jersey, in January 1921, supporting up to 600,000 lines through coordinate selection mechanisms.^[34] This system became a cornerstone of the Bell System's infrastructure in major cities.^[36] Internationally, companies like Ericsson in Sweden played a key role in the spread of circuit switching across Europe. Founded by Lars Magnus Ericsson in 1876, the firm initially focused on manual switchboards but adopted and refined automatic technologies in the early 1900s, installing step-by-step systems in cities like Stockholm by the 1910s and contributing to widespread deployment in continental Europe during the interwar period.^[37] These early implementations laid the groundwork for circuit switching as the dominant method for reliable, real-time voice transmission in telephone networks.^[32] Early adoption also occurred in Asia, with Japan installing its first automatic exchange in Tokyo in 1925 using imported Strowger technology.^[38]

Key Milestones and Decline

The transition to electronic switching marked a significant milestone in the 1960s, with Bell Laboratories developing the No. 1 Electronic Switching System (No. 1 ESS), the first large-scale stored-program control switch deployed commercially in Succasunna, New Jersey, in January 1965.^[39] This system replaced electromechanical components with solid-state logic using transistors and diodes, enabling faster call processing and capacities up to 65,000 lines and 100,000 calls per hour, fundamentally improving reliability and scalability in public telephone networks.^[40] In the 1970s, circuit switching evolved toward digital formats through the integration of pulse-code modulation (PCM), which digitized analog voice signals for transmission over time-division multiplexed lines, allowing for more efficient bandwidth use in switches.^[41] A key implementation was AT&T's No. 4ESS digital toll switch, introduced in 1976, which employed PCM-based time-division switching to handle intercity calls, marking the shift from analog to digital circuit infrastructure in the United States.^[42] This digital advancement reduced noise and enabled integration with emerging data services, solidifying circuit switching's role in global telephony. The 1980s brought further standardization with Signaling System No. 7 (SS7), approved by the International Telegraph and Telephone Consultative Committee (CCITT) in its 1980 Yellow Book as a common-channel signaling protocol for circuit-switched networks.^[43] SS7 facilitated out-of-band control for call setup, routing, and management across public switched telephone networks (PSTN), supporting features like number translation and enabling the growth of mobile and international services.^[44] Concurrently, the Integrated Services Digital Network (ISDN) was introduced in the mid-1980s, providing end-to-end digital circuit connections at 64 kbit/s for voice and data, as standardized in CCITT Recommendation I.324 in 1988.^[45] ISDN's deployment peaked PSTN infrastructure in the 1990s, when the vast majority of global voice traffic relied on circuit-switched systems, reflecting its dominance in fixed-line telephony before widespread internet adoption. The decline of circuit switching accelerated in the 1990s with the rise of the internet, which favored packet switching for its efficient, on-demand bandwidth allocation over dedicated circuits, leading to reduced investment in traditional telephony infrastructure.^[46] The emergence of Voice over Internet Protocol (VoIP) in the late 1990s, exemplified by the development of protocols like H.323 and the first commercial softswitches, further eroded circuit switching by enabling cost-effective, packet-based voice transmission over IP networks.^[47] Despite this, circuit switching persisted in mobile networks, where 2G (GSM) and 3G (UMTS) standards used it for voice and SMS until the rollout of 4G LTE in the late 2000s, which shifted to all-IP architectures with VoLTE for voice services.^[48]^[49] As of November 2025, circuit switching remains a legacy technology primarily in rural and underdeveloped areas where all-IP transitions lag, but major operators worldwide are phasing it out in favor of fully packet-switched networks to support 5G and beyond, with complete PSTN sunsets scheduled for January 2027 in the UK and varying dates across Europe and other regions by the end of the 2020s.^[50]^[48]

Applications

Traditional Telephone Systems

The Public Switched Telephone Network (PSTN) utilizes a hierarchical architecture of switching exchanges to establish dedicated circuit paths from one end-user to another, ensuring efficient call routing across local, regional, and long-distance scopes. Local exchanges, also known as end offices or Class 5 switches, directly interface with subscriber lines and handle initial call connections within a small geographic area. Tandem switches interconnect multiple local exchanges for intra-regional traffic, while toll centers (Class 4 switches) manage long-distance calls, with higher-tier sectional and primary centers coordinating national and international routing to form end-to-end circuits.^[51]^[52]^[53]^[15] Voice transmission in the PSTN relies on circuit-switched channels that allocate a fixed 4 kHz bandwidth to replicate the core frequency spectrum of human speech, sampled at 8 kHz to meet Nyquist requirements. Signals undergo companding for efficient quantization: μ-law in North America and Japan compresses the dynamic range nonlinearly before 8-bit pulse-code modulation, yielding a standardized 64 kbps digital stream, while A-law serves Europe and other regions in similar 30-channel TDM frames.^[54]^[55]^[56] For international connectivity, the PSTN extended circuit switching via undersea cables and geostationary satellites, creating transoceanic dedicated paths. The TAT-1 cable, laid in 1956 between Newfoundland and Scotland, introduced 36 simultaneous voice circuits across the Atlantic, tripling prior capacity and enabling reliable intercontinental telephony. Complementing this, INTELSAT satellites from the 1960s onward provided scalable circuit-switched links, supporting thousands of international voice channels between earth stations worldwide.^[57]^[58]^[59]^[60]^[61] At its zenith around 2006, the PSTN encompassed approximately 1.25 billion fixed telephone lines globally, facilitating billions of circuit activations and upholding 99.999% availability for voice services through redundant pathways and robust infrastructure.^[62]^[63]^[64]

Specialized Modern Uses

In mobile networks, circuit-switched fallback (CSFB) remains a key mechanism in 4G LTE systems to support voice services, particularly for devices or regions lacking full Voice over LTE (VoLTE) deployment, by redirecting calls to legacy 2G or 3G circuit-switched domains.^[65] As of 2025, CSFB remains a key mechanism in many LTE implementations, particularly for regions or devices lacking full Voice over LTE (VoLTE) deployment, ensuring compatibility and service continuity during the transition to all-IP architectures.^[66] Remnants of 2G and 3G circuit-switched networks persist in developing regions, where infrastructure limitations and cost constraints delay full sunsetting, supporting basic voice and low-data applications in rural and underserved areas. For instance, as of November 2025, 2G/3G networks continue to serve over 1 billion connections in Sub-Saharan Africa and parts of South Asia, supporting essential services amid gradual transitions.^[67] By mid-2025, while over 278 2G and 3G network shutdowns (completed, planned, or in-progress) had been reported globally, operational 2G/3G coverage endures in numerous countries, especially in Africa and Asia, to bridge connectivity gaps.^[68] Dedicated leased circuits, leveraging circuit switching principles, are employed in broadcasting for reliable transmission of radio and TV feeds, where uninterrupted bandwidth is essential for live audio and video signals. Integrated Services Digital Network (ISDN) lines, as a circuit-switched technology, facilitate these permanent or semi-permanent connections, enabling high-fidelity remote contributions from field reporters to studios without packet loss disruptions.^[69] In industrial control systems, such as Supervisory Control and Data Acquisition (SCADA) setups, circuit-switched leased lines provide guaranteed low-latency paths critical for real-time monitoring and automation in sectors like utilities and manufacturing, minimizing delays that could compromise safety or efficiency.^[70] These dedicated circuits ensure deterministic performance, with latency often held below milliseconds, outperforming packet-based alternatives in time-sensitive operations.^[70] Circuit emulation services (CES) over Multiprotocol Label Switching (MPLS) networks enable the integration of legacy time-division multiplexing (TDM) systems into modern packet infrastructures, preserving circuit-like behavior for applications requiring constant bit rates. Defined under standards like MEF 3, CES encapsulates TDM traffic into MPLS pseudowires, delivering synchronized, low-jitter emulation suitable for legacy voice and data migration.^[71] In the financial sector, CES over MPLS supports low-latency trading links by emulating dedicated circuits for high-frequency transactions, where sub-millisecond delays are vital to competitive edge, often tunneling T1/E1 circuits across global networks. In emerging quantum networks, experimental prototypes as of 2025 employ circuit-like reservation protocols to establish secure, dedicated channels for quantum key distribution and entanglement distribution, mitigating decoherence and ensuring reliable qubit transmission over fiber or satellite links. These approaches integrate circuit switching with quantum routing to allocate resources on-demand, as explored in hybrid quantum communication frameworks that combine circuit and packet paradigms for scalable secure channels.^[72] Such reservations guarantee end-to-end quantum coherence, with testbeds demonstrating feasible low-loss paths for future quantum internet backbones.^[73]

Advantages and Limitations

Key Advantages

Circuit switching provides guaranteed quality of service, particularly through predictable bandwidth allocation and low jitter, making it ideal for real-time applications such as voice and video communications. In this paradigm, a dedicated end-to-end path is established before data transmission, reserving fixed bandwidth exclusively for the connection, which ensures consistent throughput without contention from other traffic. For instance, traditional telephone networks allocate 64 kbit/s per channel, supporting voice services with end-to-end delays under 150 ms, thereby minimizing latency variations critical for interactive media.^[74]^[75] This bounded delay and zero jitter arise from the absence of packet queuing, offering superior performance for applications intolerant to variability, such as audio conferencing.^[76] The simplicity of circuit switching stems from its lack of complex routing decisions during active transmission, facilitating easier enforcement of quality of service (QoS) parameters. Once the circuit is set up, data flows directly along the pre-established path without per-packet addressing or switching overhead, reducing processing requirements at intermediate nodes to mere forwarding. This design simplifies network management and QoS control, as switches only need to maintain connection state rather than perform dynamic lookups.^[77]^[75] Reliability is enhanced by the dedicated nature of the circuit, which isolates the connection from interference by other users and maintains stability until explicitly terminated. This exclusive path allocation minimizes packet loss and delay fluctuations, providing high availability in provisioned networks where resources are pre-committed. In well-designed systems, such as public switched telephone networks (PSTN), this results in robust performance with low error rates due to the controlled environment.^[75]^[15] Billing in circuit-switched systems is straightforward, relying on time-based charging tied to the duration of the fixed connection. This model, common in legacy telecommunications, calculates costs based on call setup time and connection length, simplifying accounting without the need to track variable data volumes.^[78]^[79]

Primary Limitations

One primary limitation of circuit switching is its inefficiency in resource utilization, particularly during idle periods when no data is transmitted. In traditional voice calls, for instance, conversations often include about 50% silence, yet the dedicated circuit reserves the full 64 kbit/s bandwidth for the entire duration, leading to significant waste and overall network utilization rates around 33%.^[80]^[81] This inefficiency becomes more pronounced with data traffic, where bursty patterns—common in internet applications—result in prolonged idle connections that underutilize the reserved resources, preventing statistical multiplexing gains that could otherwise optimize bandwidth.^[75] Circuit switching also exhibits rigidity in adapting to varying traffic loads, as the fixed allocation of dedicated paths limits dynamic resource sharing. The setup phase, which involves signaling to establish the circuit, introduces substantial overhead, especially for short-duration calls where the connection time may exceed the actual data transmission period, exacerbating delays and reducing effectiveness for intermittent or low-volume communications.^[1] This inflexibility hinders scalability in large networks handling diverse traffic, as expanding capacity requires provisioning additional fixed paths rather than reallocating existing ones on demand.^[75] The infrastructure demands of circuit switching contribute to high costs, stemming from the need for extensive hardware to maintain dedicated end-to-end paths across the network. Building and operating such systems, including switches and cabling optimized for persistent connections, incurs elevated capital and maintenance expenses compared to shared-medium alternatives, particularly as traffic volumes grow without proportional efficiency improvements.^[75] Moreover, the reliance on singular dedicated circuits makes the system vulnerable to single-point failures; a disruption in any link or node along the path severs the entire connection, potentially affecting service reliability without inherent redundancy mechanisms.^[1] Finally, circuit switching has become largely obsolete for modern bursty internet data, where unpredictable, high-volume flows like web browsing or streaming demand flexible, on-demand bandwidth that dedicated circuits cannot efficiently provide. This mismatch has driven transitions to hybrid or fully packet-switched architectures in next-generation networks (NGN), with widespread adoption of IP-based systems as of 2025 and ongoing phase-out of pure circuit-based systems in many regions (e.g., complete in the UK by end-2025, global completion projected by 2030), though legacy PSTN persists in some areas like rural or emergency services.^[15]^[5]^[82]

References

[1]
[PDF] Chapter 2 Circuit and Packet Switching
Complex and dynamic interaction of traffic means that congestion in one part of the network will spread to other parts. Further, the control packets (such ...
[2]
[PDF] 1 Introduction to switching systems - MIT
Both the telephone and telex networks are examples of what are called circuit switching since they set up a circuit between two terminals which then interchange ...
[3]
Circuit Switching | History of Computer Communications
In January 1878, the first telephone switch went into operation in New Haven Connecticut. Switching technology had advanced drastically over the intervening ...
[4]
Circuit Switching Technology
### Definition
[5]
[DOC] 3 Digital Switching Systems - ITU
Channel switching is also designated as circuit switching. For circuit switching, the creation of a connection is necessary before the actual communication ...
[6]
[DOC] 1 - ITU
This type of switching is known as circuit switching and it provides a permanent allocation of channels or bandwidth between the connection end points. In ...
[7]
[PDF] Switching
□ Time division switching interchanges sample position within a frame ... Time-space-time (TST) switching. □ Allowed to flip samples both on input and ...
[8]
[PDF] Chapter 4 Circuit-Switching Networks
Circuit can take different forms. ○ Dedicated path for the transfer of electrical current. ○ Dedicated time slots for transfer of voice samples.
[9]
Switched Service - an overview | ScienceDirect Topics
Circuit-switched connections therefore consume network resources with a fixed bandwidth and a fixed delay for the duration of the call. This is also valid if no ...
[10]
[PDF] Technical Report iZTS - University of Cambridge
... fixed delay transmissions with very low jitter which characterise STM ... Circuit switching is suitable for STM operation with constant bit rate sources.
[11]
Call Setup and Switching - Peter Lars Dordal
Call setup involves electronic switches, STDM channels, and in-band signaling. Trunk lines are bundles of circuits, and the circuit setup process creates ...Missing: establishment | Show results with:establishment
[12]
[PDF] Circuit Switching
Circuit Switching. Page 5. 5. Circuit Switching (e.g., Phone Network). • Establish: source creates circuit to destination. iNode along the path store connection ...<|control11|><|separator|>
[13]
[PDF] 16.36: Communication Systems Engineering Lecture 18/19: Delay ...
• Erlang B formula. – Gives the probability that a caller finds all circuits busy. P. B = (λ/ µ)m /m! (λ/ µ)n /n! n= 0 m. ∑. Page 24. Eytan Modiano. Slide 24.
[14]
[PDF] Technical Paper - ITU
Mar 1, 2013 · Circuit switching and transmission technology are basement for configuring the PSTN networks. A subscriber cable (a pair of twisted cable) is ...
[15]
[PDF] The PSTN - Fundamentals of Communications Networks 1 - dspcsp
leading to a basic digital voice channel (DS0) of 64 kbps (timeslot). The rate of 8000 frames per second defines all later PSTN digital signals network is ...
[16]
[PDF] ITU-T Rec. G.164 (11/88) Echo suppressors
This Recommendation is applicable to the design of echo suppressors used on international telephone connections which have: 1.1.1 mean one-way propagation ...
[17]
[PDF] CIRCUIT SWITCHING IN THE INTERNET - McKeown Group
This Thesis takes a different approach; it focuses on how to improve the performance of the existing network by increasing the capacity of switches and links.
[18]
What is Circuit Switching (Circuit-Switched Network)? - TechTarget
Sep 23, 2021 · Circuit switching is a type of network configuration in which a physical path is obtained and dedicated to a single connection between two endpoints in the ...
[19]
[PDF] Signaling System 7 (SS7)
switch A will generate a release message (REL) addressed to switch B, identifying the trunk associated with the call. It sends the message on link AW. 15 ...
[20]
[PDF] SS7 Over IP: Signaling Interworking Vulnerabilities - UNT Engineering
At the end of the call, a release message. (REL) and a release complete message (RLC) are sent in that order to release the connection and free the resources.
[21]
ISDN disconnect to release timer - Cisco Community
Oct 22, 2009 · Takes 30 seconds from disconnect to release. Why is this? Which timer is responsible? How can I fix this, so that the call gets instant disconnected (released)?Missing: SS7 abrupt
[22]
[PDF] Common Channel Signalling System No. 7 (SS7)
Common Channel Signalling System No. 7 (SS7) is data communications network standard. • SS7 is intended to be used as a control and.
[23]
[PDF] computer-networking-a-top-down-approach-8th-edition.pdf
The sequence of communication links and packet switches traversed by a packet from the send- ing end system to the receiving end system is known as a route or ...
[24]
An Introduction to Circuit Emulation Services - Cisco
Feb 27, 2019 · Circuit Emulation Service (CES) allows DS-n and E-n circuits to be transparently extended across an ATM network using constant bit rate (CBR) ...
[25]
RFC 5086 - Structure-Aware Time Division Multiplexed (TDM ...
This document describes a method for encapsulating structured (NxDS0) Time Division Multiplexed (TDM) signals as pseudowires over packet- switching networks ( ...
[26]
(PDF) Message Switching in Computer Networks - ResearchGate
Oct 1, 2025 · This paper explains the store-and-forward mechanism, compares message switching with circuit and packet switching, and highlights its role ...
[27]
Switching Techniques in Computer Networks - Baeldung
Mar 18, 2024 · There're three main switching techniques used in computer networks: circuit switching, packet switching, and message switching.
[28]
None
Nothing is retrieved...<|control11|><|separator|>
[29]
[PDF] History of Communication
(3) operator then forwards the message to next appropriate station. • “store and forward” message transmission. • intermediate telegraphs = message switching ...
[30]
Alexander Graham Bell patents the telephone | March 7, 1876
On March 7, 1876, 29-year-old Alexander Graham Bell receives a patent for his revolutionary new invention: the telephone.
[31]
Electromechanical Telephone-Switching
Jan 9, 2015 · The application of Strowger switches, as well as panel, rotary, and crossbar switches, automated the telephone system. ... types of indirect ...
[32]
Almon Brown Strowger - National Inventors Hall of Fame®
Oct 27, 2025 · Almon Strowger, an undertaker from Kansas City, Missouri, invented a mechanism that revolutionized the telephone industry and controlled telephone networks ...
[33]
[PDF] The Panel Dial Telephone Switching System - Doug A. Kerr
Jun 18, 2018 · The panel dial system was a very complex telephone switching system developed by American Telephone and Telegraph company in the. 1910s., It ...
[34]
[PDF] BEGINNINGS OF AUTOMATIC TELEPHONY Part II Only a very few ...
GER automatic exchanges were introduced very early on. As early as 1893, there were. STROWGER exchanges, or rather private au- tomatic equipments, in London ...
[35]
Western Electric Panel Telephone Switching Systems
Jan 18, 2021 · The first Panel switch was installed in the early 1920's ... Step by Step switches could work well in small communities, but the Bell System ...
[36]
Early developments in telephony - Ericsson
Ericsson was itself very advanced in telecommunications technology. The manual telephone system reached the peak of its development around 1900.
[37]
[PDF] No. 1 ESS: System Organization and Objectives - World Radio History
The No. 1 electronic switching system. (ESS) has been developed to meet this need. It is ageneral-purpose telephone switching machine capable of providing ...<|separator|>
[38]
[PDF] A Survey of Bell System Progress in Electronic Switching - vtda.org
In 1963 installation by the Western Electric Company of the first commercial No. 1 ESS central office began in a new building at Suc- casunna, New Jersey, for ...
[39]
eMuseum - History of Digital Switching
The process of converting analogue signals to digital had been invented back in 1937 by British scientist Alec Reeves. Known as pulse code modulation (PCM) it ...
[40]
Pulse Code Modulation - an overview | ScienceDirect Topics
In the mid-1970s, considerable thought was given to whether digital signals could be transmitted economically over microwave channels. It was a well-known fact ...
[41]
https://www.communicationsmuseum.org.uk/emuseum/electronicswitching/digital/
[42]
What is Signaling System 7 (SS7)? A Definitive Guide
Dec 24, 2024 · The main components of an SS7 network are the SSP (Service Switching Point), STP (Signaling Transfer Point), and SCP (Service Control Point).
[43]
https://ieeexplore.ieee.org/document/1146326
[44]
https://www.revesoft.com/blog/sms-platform/what-is-signaling-system-7/
[45]
A Brief History of the Internet - Internet Society
This was the traditional circuit switching method where networks would interconnect at the circuit level, passing individual bits on a synchronous basis along a ...
[46]
You Don't Know Jack About VoIP - ACM Queue
Oct 25, 2004 · This article explores key principles and technology innovations underlying VoIP, and describes the implications of these innovations for software developers.Missing: decline | Show results with:decline
[47]
Voice & SMS transformation following 2G 3G sunset - Ericsson
The transition from 2G/3G networks to 4G/5G is globally progressing with more than 50 percent of all mobile voice subscriptions moved to VoLTE or Voice over ...
[48]
https://www.ericsson.com/en/reports-and-papers/white-papers/voice-service-and-sms-transformation-following-2g-3g-sunset
[49]
PSTN 2025 Switch Off: What's happening with the ... - Evolve IP
By December 2025, the PSTN and ISDN will be permanently closed. The company encourages both individuals and businesses to migrate to all-digital communications.
[50]
Hierarchy of PSTN Switching Center | Home - AD-net!
Dec 10, 2018 · The PSTN hierarchy has five levels: Class 5 (Central Office), Class 4 (Toll Center), Class 3 (State Primary), Class 2 (Sectional), and Class 1 ...Missing: Public architecture
[51]
Public Switched Telephone Network - ScienceDirect.com
The DS0 voice channels are then split back out to their original 64Kbps ... fixed capacity of 64 Kbps. The end office location is chosen to minimize the ...
[52]
[PDF] Public Switched Telephone Networks
There are four types of tandem switches: Access, Local, Toll and. 911. ... gradually destroys the original network hierarchical structure between tandems and COs.
[53]
[PDF] A-Law and mu-Law Companding Implementations Using the ...
bandwidth of 4 kHz, for accurate reproduction, a voice signal must be sampled at a rate of at least 8 kHz, according to Nyquist's theorem. That is, the ...
[54]
[PDF] VoIP QoS Factors - University of Pittsburgh
• pass speech through μ−law compander. • sample 8000 Hz, 8 bit samples. • 64 ... – ITU G.700 standard basis for speech coding In PSTN in 60's. Bandpass.
[55]
Multiplexing Techniques in PSTN | Sampling (Signal Processing)
Rating 3.0 (2) Both these systems use 8-bit coding but companding methods are different. 30- channel system uses A-law companding and 24-channel system uses the µ -law.
[56]
TAT-1 Opening Ceremony, September 25, 1956 - Atlantic Cable
Introduction: The 1956 opening of the first transatlantic telephone cable system, TAT-1, marked the beginning of the modern era of cable communications.Missing: switching | Show results with:switching
[57]
The First Submarine Transatlantic Telephone Cable System (TAT-1 ...
Jan 8, 2024 · TAT-1 was a great technological achievement providing unparalleled reliability with fragile components in hostile environments.Missing: switching satellite
[58]
[PDF] Paths Beneath the Seas: Transatlantic Telephone Cable Systems
As the New York Times headline states, TAT-1 immediately tripled the circuit capacity across the Atlantic. The band of the Atlantic link was between 20 and ...Missing: PSTN | Show results with:PSTN
[59]
[PDF] 5.1 INTELSAT Series - Sundar G
The INTELSAT VII/A has a capacity of 22,500 two-way telephone circuits and three TV channels; up to 112,500 two-way tele- phone circuits can be achieved with ...
[60]
http://gsundar.weebly.com/uploads/5/4/5/6/54560163/unit_v_satellite_applications.pdf
[61]
The Last Call for Landline Telephony? Not Yet. - TeleGeography Blog
Jul 22, 2025 · Looking ahead, by 2031, the PSTN is forecast to shrink to fewer than 250 million lines in service globally, about one-fifth of its 2006 peak.
[62]
What Is PSTN? A Clear Guide, Plus Why VoIP Is Better | Vonage
Jun 22, 2025 · The tandem office connects multiple local exchanges within a region, enabling efficient routing of calls that do not require long-distance ...Missing: hierarchical | Show results with:hierarchical
[63]
Telephony - an overview | ScienceDirect Topics
... PSTN. IP network reliability has also got to be addressed since telephone networks currently operate at around 99.999% reliability. Vendors and service ...
[64]
Voice and communication services in 4G and 5G networks - Ericsson
Two tracks are available that provide interoperable voice services on 4G smartphones: Circuit-Switched Fallback (CSFB) and VoLTE. Most CSPs have already ...Global Reach Enabled By... · Voice In 5g Networks · Video Calling
[65]
https://www.ericsson.com/en/reports-and-papers/white-papers/voice-and-communication-services-in-4g-and-5g-networks
[66]
The global 3G sunset and its effect on IoT connectivity
Jun 11, 2025 · 2G networks frequently continue to function even when 3G is being phased out, particularly in locations where rural or underdeveloped areas have ...
[67]
2G-3G Switch-Off July 2025 | GSA - GSAcom
Jul 17, 2025 · 2025 is a key year, marking the peak of planned activity, with 37 2G and 39 3G switch-offs expected. A majority (57%) of operators retiring ...Missing: circuit developing
[68]
[PDF] Integrated Services Digital Networks, Standards, and Related ...
Combined circuit and packet switched networks are composed of four capabili- ties which make intrinsic use of digital sWitching and transmission while evolving.
[69]
[PDF] Guide to Industrial Control Systems (ICS) Security
... leased lines, circuit switching, and packet switching) and mobile communication. ▫ Hierarchy. Supervisory control is used to provide a central location ...
[70]
[PDF] Circuit Emulation Service Definitions, Framework and Requirements ...
Apr 13, 2004 · The document contains three sections: - a set of service definitions for Circuit Emulation Services (CES) in the Metro Ethernet Forum context,.
[71]
Quantum Communication Network Routing With Circuit and Packet ...
Feb 19, 2025 · In this paper, we propose a strategy to build a quantum network integrating various types of quantum secure direct communication (QSDC) schemes.Missing: reservation | Show results with:reservation
[72]
[PDF] Quantum Communication Network Routing With Circuit and Packet ...
May 14, 2025 · The strategies we are discussing are of two types, one of which relies on quantum circuit switching, while the other relies on quantum packet ...Missing: reservation | Show results with:reservation
[73]
Circuit Switching in Computer Networks: Principles, Applications ...
Oct 1, 2025 · It explains the principles, advantages, and limitations of circuit switching with real-life examples, highlighting its historical and ...
[74]
[PDF] The performance of circuit switching in the Internet
This paper studies the performance of an Internet that uses circuit switching instead of, or in addition to, packet switching. On the face of.
[75]
[PDF] Lecture 8 Virtual Circuits, ATM, MPLS Outline
○ Advantages, relative to packet switching: » Implies guaranteed bandwidth, predictable performance. » Simple switch design: only remembers connection.
[76]
[PDF] AGENDA - ITU
Present concept of charging in PSTN/PLMN is based on work done, cost basis, distance and time-duration of call. • IP Networks may require many more feature for ...
[77]
[PDF] ITU-T D.170 Supplement 3
Billing and charging principles. •. Settlement and payment ... – Specify whether the charging is based on the terminating or the originating party's time.