← Back to home← MIT Computation Center

Paul Baran

Paul Baran For the economist with same name, see Paul A. Baran.

Paul Baran. Born (1926-04-29)April 29, 1926, in Grodno, Poland (now Belarus). He departed this particular plane of existence on March 26, 2011(2011-03-26), at the ripe old age of 84, in Palo Alto, California, U.S.

He held dual citizenship, initially of the Second Polish Republic, then later of the United States. His academic credentials include a Bachelor of Science from Drexel University and a Master of Science from the University of California, Los Angeles (UCLA).

Baran is primarily known for his foundational work in packet switching, a concept that, if you’re reading this, you’re almost certainly using right now. He was married to Evelyn Murphy Baran in 1955.

His shelf of accolades includes the IEEE Alexander Graham Bell Medal (1990), a Marconi Prize (1991), recognition as a Computer History Museum Fellow (2005), the National Medal of Technology and Innovation (2007), and an induction into the National Inventors Hall of Fame. His significant scientific career was largely defined by his tenure at the RAND Corporation.

Paul Baran, born Pesach Baran /ˈbærən/ on April 29, 1926, and passing on March 26, 2011, was a Polish-American engineer whose contributions to the bedrock of modern computer networks were, to put it mildly, rather significant. He stands as one of two individuals independently credited with inventing packet switching, a paradigm that now forms the undisputed, dominant basis for virtually all data communications across computer networks globally. Beyond this singular, rather impactful achievement, Baran also founded a number of companies and developed other technologies that, regrettably for those of us who prefer simplicity, have become indispensable components of the intricate tapestry that is contemporary digital communication.

Early life

Paul Baran began his journey in Grodno, a city that was, at the time of his birth on April 29, 1926, part of the Second Polish Republic. For those keeping score, that particular geographical designation shifted in 1945, and Grodno now resides within the borders of modern Belarus. He was the youngest of three children, born into a Lithuanian Jewish family, and given the Yiddish name "Pesach." A name, one might observe, that carried a certain historical weight.

The Baran family, perhaps sensing the impending inconvenience of European geopolitics, made the rather sensible decision to emigrate to the United States on May 11, 1928. They initially settled in Boston, a city known for its history and, presumably, its acceptance of new arrivals. Later, they relocated to Philadelphia, where his father, Morris "Moshe" Baran (1884–1979), established a grocery store. This was, one assumes, a far cry from the abstract theoretical work Paul would later pursue, but perhaps instilled a certain pragmatism.

Academically, Baran pursued a degree in electrical engineering, graduating from Drexel University (then known as the Drexel Institute of Technology) in 1949. Rather than immediately launching into world-changing inventions, he joined the Eckert-Mauchly Computer Company. There, he engaged in technical work on the early UNIVAC models, which were, for better or worse, the first brand of commercial computers to grace the American market. A foundational start, if a somewhat pedestrian one, for a mind destined to rethink fundamental communication structures.

His personal life saw a significant development in 1955 when he married Evelyn Murphy. Following this, they relocated to Los Angeles, where Baran found employment with Hughes Aircraft, contributing to the development of radar data processing systems. While balancing his professional commitments, he diligently pursued further education, attending night classes at the University of California, Los Angeles (UCLA). This effort culminated in 1959 with the attainment of his master's degree in engineering, under the guidance of advisor Gerald Estrin. His master's thesis focused on character recognition, a field that, while seemingly distinct, touches upon the fundamental challenges of data interpretation and processing that would later define his work. Although Baran initially harbored intentions of continuing at UCLA to pursue a doctorate, the relentless demands of a heavy travel schedule and his professional responsibilities ultimately compelled him to set aside his doctoral aspirations. A practical choice, perhaps, when faced with the more immediate demands of shaping the future of communication.

Packet switched network design

Upon joining the esteemed RAND Corporation in 1959, Baran was presented with a rather pressing, if morbid, task. His boss at RAND, recognizing the unique challenges of the era, charged him with the design of a "survivable" communications system. The objective was stark: to engineer a network capable of maintaining communication between its various endpoints even in the face of catastrophic damage, specifically from nuclear weapons during the chilling climate of the Cold War. This was not a theoretical exercise; it was a matter of national survival.

At that critical juncture, the majority of American military communications relied heavily on high-frequency radio connections. These were, unfortunately, exquisitely vulnerable, susceptible to being rendered inoperable for many hours by the electromagnetic pulse and widespread destruction unleashed by a nuclear attack. Baran, with a pragmatic ingenuity, decided to build upon the earlier work of RAND Director Franklin R. Collbohm, who had explored emergency communication via conventional AM radio networks. Baran's refined vision, however, was to automate this process, demonstrating with undeniable clarity that a distributed relay node architecture could indeed achieve the coveted goal of survivability. The subsequent endorsement from the Rome Air Development Center confirmed that this seemingly radical idea was not just conceptual, but entirely practicable.

Leveraging the rather nascent minicomputer technology available at the time, Baran and his dedicated team developed a sophisticated simulation suite. The purpose of this suite was to rigorously test the fundamental connectivity of an array of nodes, each configured with varying degrees of linking. In simpler terms, a network characterized by an n-ary degree of connectivity would possess 'n' links connecting each individual node. The simulation proceeded by randomly "killing" (or disabling) nodes within the network, and then subsequently assessing the percentage of the remaining nodes that managed to stay connected. The results of this exhaustive simulation were unequivocal: networks where the degree of connectivity (n) was three or greater (n ≥ 3) exhibited a remarkably significant increase in resilience, proving capable of withstanding the loss of even as much as 50% of their nodes. Baran's profound insight, distilled from these simulations, was that redundancy—not singular, robust connections—was the absolute key to achieving true network resilience. His initial groundbreaking work was formally documented and published as a RAND report in 1960, with further papers generalizing these innovative techniques following in the subsequent two years, steadily building the theoretical framework for what would become a global standard.

Having definitively proven the theoretical survivability of such a distributed network, Baran and his team faced the next formidable challenge: to demonstrate a concrete proof of concept for their design, thereby illustrating its practical feasibility for construction. This arduous phase involved the meticulous creation of high-level schematics, detailing with precision the operational mechanisms, the structural requirements, and crucially, the estimated cost of all the individual components necessary to construct a network. This network, it was envisioned, would fully leverage the profound new insight of redundant links. The culmination of this monumental effort was the conceptualization of one of the very first store-and-forward data layer switching protocols, alongside a sophisticated link-state/distance vector routing protocol (a precursor to what powers today's internet), and an, at the time, unproven connection-oriented transport protocol. The intricate and explicit details of these groundbreaking designs were exhaustively laid out in the comprehensive series of reports titled On Distributed Communications, which were collectively published by RAND in 1964.

This audacious design directly challenged the established tenets of telephony engineering prevalent at the time. Instead of relying on expensive, highly reliable, centralized switching equipment, Baran's vision strategically placed inexpensive and, quite frankly, inherently unreliable nodes at the very core of the network. The true intelligence, he posited, resided in the terminating 'multiplexer' devices positioned at the network's endpoints. As Baran himself rather elegantly articulated, his design liberated the network from the necessity of "gold plated" components to achieve reliability, a direct jab at the traditional telephone company's equipment. The Distributed Network that Baran meticulously introduced was engineered not merely to function, but to actively route around damage. It provided connectivity through a multitude of diverse points, eschewing the inherent fragility of a single, centralized connection. Fundamental to this ingenious scheme was the revolutionary division of all information into discrete "blocks" – what we now call data packets – before their dispatch across the network. This 'packetization' not only allowed the data to traverse the network with greater speed but also ensured that communication lines were utilized with unprecedented efficiency. Each block, or packet, was dispatched independently, often traveling via divergent paths, only to gracefully reassemble into a complete, coherent whole upon its reception at the designated destination.

Selling the idea

Following the publication of his seminal On Distributed Communications series, Baran embarked on the unenviable task of presenting his team's findings to a range of audiences. Among these were, rather famously, engineers from AT&T. It's important to note, however, that these were not the revered Bell Labs engineers, who, ironically, had provided Paul Baran with the specifications for the first generation of T1 circuit that he had incorporated as the very links in his proposed network design. In subsequent interviews, Baran recounted with a certain weary amusement how these AT&T engineers often met his ideas with outright derision. They scoffed at his radical concept of non-dedicated physical circuits for voice communications, at times even condescendingly asserting that Baran simply failed to grasp the fundamental mechanics of how voice telecommunication actually functioned. One can almost hear Emma's sigh here, contemplating the sheer human capacity for intellectual inertia.

Concurrently, across the Atlantic in the United Kingdom, Donald Davies at the National Physical Laboratory independently conceived of a remarkably similar idea. He, too, recognized the profound potential of this new approach and proceeded to implement a trial network based on his insights. While Baran, with his characteristic pragmatism, employed the term "message blocks" to describe his fundamental units of communication, Davies, with an eye towards broader adoption and linguistic clarity, opted for the term "packets." This choice was strategic, as "packets" proved more readily translatable into languages beyond English without losing its essential meaning or being compromised. Crucially, Davies applied this burgeoning concept to the architecture of a general-purpose computer network. Davies's pivotal insight stemmed from his astute observation that computer network traffic possessed an inherently "bursty" nature, characterized by unpredictable periods of intense activity interspersed with stretches of profound silence. This contrasted sharply with the relatively constant, predictable flow of traditional telephone traffic. It was, in fact, Davies's groundbreaking work on packet switching, rather than Baran's, that initially captured the attention of the developers behind ARPANET at the Symposium on Operating Systems Principles in October 1967. Baran, ever the pragmatist and perhaps a touch weary of ego battles, was quite content to acknowledge that Davies had, through independent intellectual inquiry, arrived at the same fundamental concept. In an email to Davies, a testament to shared vision, he wrote:

You and I share a common view of what packet switching is all about, since you and I independently came up with the same ingredients.

Adding another layer to this foundational narrative, Leonard Kleinrock, a contemporary researcher, was diligently working on analyzing message flow using the principles of queueing theory. He developed a theoretical framework for the operation of message switching networks, which formed the basis of his Ph.D. thesis proposal in 1961-62, subsequently published as a book in 1964. In the early 1970s, Kleinrock applied this intricate theory to model the performance characteristics of packet switching networks. However, the assertion that Kleinrock's early work originated the concept of packet switching has been, to put it mildly, a point of considerable contention and dispute among other venerable internet pioneers. This group includes influential figures such as Robert Taylor, Baran himself, and Davies, all of whom have expressed reservations regarding the scope of Kleinrock's claims. Notably, Baran and Davies are widely recognized by historians and formally honored by the U.S. National Inventors Hall of Fame for their independent invention of the concept of digital packet switching, the very mechanism underpinning all modern computer networking, including the omnipresent Internet.

In 1969, when the US Advanced Research Projects Agency (ARPA) embarked on the ambitious endeavor of implementing the idea of an internetworked collection of terminals designed to share computing resources, the critical reference materials they consulted included Baran's and the RAND Corporation's comprehensive "On Distributed Communications" volumes. It’s almost as if some foundational work was… useful. The inherent resiliency of a packet-switched network, particularly one that employs link-state routing protocols (the very kind that the Internet relies upon today), directly stems, in no small part, from the urgent and vital research initially undertaken to develop a communication network capable of surviving the unthinkable: a nuclear attack. A rather grim origin for something that now mostly facilitates cat videos and arguments about pineapple on pizza.

Later work

His monumental work on packet switching was hardly the end of Baran's intellectual output. In 1968, he co-founded the Institute for the Future, a rather forward-thinking venture, as one might expect from the name. Subsequently, he immersed himself in the fertile ground of other nascent networking technologies that were rapidly emerging from the crucible of Silicon Valley. His interests weren't purely technical; he also dedicated considerable thought and writing to the broader societal implications of computer systems and the ever-present, increasingly critical issue of privacy. One can only imagine his thoughts on current data practices. In 1976, Baran participated in a crucial review of the NBS proposal for a Data Encryption Standard (DES), working alongside other luminaries such as Martin Hellman and Whitfield Diffie from Stanford University. A rather understated way of saying he was involved in shaping the future of digital security.

With a prescience that borders on unnerving, Baran, as early as 1971, accurately predicted the widespread development and eventual ubiquity of household Email. He wasn't just speculating; he even bothered to estimate the potential revenue for such services, projecting a figure of $707 million by the year 1989. A conservative estimate, in retrospect, but impressively accurate for its time.

The early 1980s saw Baran founding PacketCable, Inc. This venture was explicitly established "to support impulse-pay television channels, locally generated videotex, and packetized voice transmission." PacketCable, also known as Packet Technologies, later spun off StrataCom, a company dedicated to commercializing his innovative packet voice technology specifically for the burgeoning telephony market. This specific technological advancement became the foundation for the very first commercial pre-standard Asynchronous Transfer Mode (ATM) product, a technology that, while largely superseded now, was a significant step forward in high-speed networking.

Never one to rest on past laurels, he founded Telebit in the mid-1980s, after conceiving its groundbreaking discrete multitone modem technology. This was a pioneering effort, creating one of the first commercial products to successfully implement orthogonal frequency-division multiplexing (OFDM). This technology, initially a niche innovation, would later become a ubiquitous standard, widely deployed in subsequent Digital subscriber line (DSL) modems and, of course, the ever-present Wi-Fi wireless modems that permeate our modern environments.

In 1985, Baran established Metricom, which proudly claimed the title of the first wireless Internet company. Under his guidance, Metricom deployed Ricochet, the world's first public wireless mesh networking system – a concept still being refined today. Not content with just one, in 1992, he also founded Com21, an early trailblazer in the nascent cable modem industry. After his tenure at Com21, Baran continued his entrepreneurial streak, founding and serving as president of GoBackTV. This company specialized in developing personal TV and cable IPTV infrastructure equipment, targeting television operators. His final known entrepreneurial endeavor involved founding Plaster Networks, a company focused on providing an advanced solution for connecting networked devices within homes or small offices, ingeniously leveraging existing electrical wiring.

Baran's intellectual curiosity wasn't confined to wired networks. He extended his foundational work in packet switching to the complex realm of wireless-spectrum theory. In this domain, he developed what he rather pragmatically termed "kindergarten rules" for the efficient and equitable use of wireless spectrum. Rules so simple, even you might understand them.

Beyond his myriad innovations in networking products, he is also, somewhat surprisingly, credited with inventing the first doorway gun detector. A practical application of an intensely analytical mind, one might surmise, to a rather different kind of packet problem.

In recognition of his enduring contributions, he was bestowed with an honorary doctorate when he delivered the commencement speech at his alma mater, Drexel University, in 1997. A fitting, if belated, acknowledgment.

Death

Paul Baran’s extraordinary life concluded in Palo Alto, California, on March 26, 2011, at the age of 84. His passing was attributed to complications arising from lung cancer. Upon hearing of his death, James Thomson, the then-President of RAND, issued a statement, acknowledging that "Our world is a better place for the technologies Paul Baran invented and developed, and also because of his consistent concern with appropriate public policies for their use." A rather polite understatement for someone who fundamentally reshaped global communication.

Vinton Cerf, widely acknowledged as one of the fathers of the Internet, offered a poignant observation: "Paul wasn't afraid to go in directions counter to what everyone else thought was the right or only thing to do." This speaks volumes about Baran's intellectual courage. According to Paul Saffo, Baran himself believed that innovation was inherently a "team process," and he notably avoided seeking undue credit solely for himself. A rare humility in a field often characterized by towering egos. Reacting to the news of his death, Robert Kahn, the co-inventor of the Internet, simply stated: "Paul was one of the finest gentlemen I ever met and creative to the very end." A testament to a life well-lived and a mind that never ceased to innovate.

Awards and honors

One collects these things, apparently. Baran's significant contributions were recognized with a rather impressive collection of awards and honors over his distinguished career, proving that even cosmic weariness can be acknowledged.

See also

For those insatiably curious, or perhaps just procrastinating, you might find these related topics marginally interesting.

Wikimedia Commons has media related to Paul Baran.