The Electrochemical Society Interface • Winter 2017 • www.electrochem.org 71 Emergence of 5G F rom connected thermostats to self-driving vehicles, advanced connectivity is transforming our lives while saving energy and reducing pollution, thus making living more sustainable. This is the hallmark of the Internet of things (IoT). But the full extent of IoT possibilities is yet to be realized. One of the key foundational technologies that will serve as the core architecture for future capabilities is 5G—the 5th generation telecommunication systems. Unlike its predecessors, 5G is going to be a new heterogeneous ecosystem that includes integration of 4G, Wi-Fi, millimeter waves, and other wireless access technologies, utilizing bandwidths from various frequency bands ranging from <4 GHz to 100 GHz. 1 Different usage scenarios demand different capabilities that can be attained in different frequency bands. As per the International Telecommunication Union (ITU), three main usage scenarios are envisaged to expand and support a diverse array of applications for International Mobile Telecommunications (IMT) 2020 and beyond, namely: 1) enhanced mobile broadband; 2) massive machine-type communications; and 3) ultra-reliable and low latency communications (as illustrated in Fig. 1), along with the key performance indicators (KPIs) associated with them. 2 The taming of this spectrum is considered essential to achieving the 5G vision of a truly connected world. RF CMOS: Requirements and Key Challenges The focus of this article is mainly on two of the proposed frequency bands, centered around 6 GHz and 28 GHz, which are designated for connected vehicles and indoor hotspot applications, respectively. Demanding KPIs of various applications call for innovation in radio frequency integrated circuit (RFIC) design. For example, for the connected vehicles in urban grids and highways, high mobility (>300 km/h), high reliability and low latency (≤1 ms) are essential. Lower latency can be achieved by bringing circuit components in a mixed-signal chip closer to one another. However, the physical proximity of the electromagnetic (EM) noise aggressor, viz., digital components, and the victim, viz., analog components, at such high frequencies may lead to serious electromagnetic interference which, in turn, may degrade the reliability of the devices and the systems. 3,4 This is of serious concern in designing the RF CMOS circuit layout that uses all frequency bands related to 5G. Therefore, to continue miniaturization of radio frequency integrated circuits while maintaining high reliability, either a novel circuit design approach should be adopted or an on-chip EM noise suppressor must be used. Besides electromagnetic compatibility issues, progress in bringing analog, digital, and RF circuits on one chip—commonly known as system-on-chip (SoC)—is hindered greatly Magnetic Nanoferrites for RF CMOS: Enabling 5G and Beyond by Ranajit Sai, S. A. Shivashankar, Masahiro Yamaguchi, and Navakanta Bhat due to a crucial passive component—the inductor. Inductors are widely used in RF integrated circuits. They are essential components of low-noise amplifers (LNA), power amplifers (PA), oscillators, flters, etc. But, often, they are the largest components in the circuit, costing more than 50% of chip-area. Apart from “chip real estate,” their large footprint results in a large parasitic capacitance at high frequencies which, in turn, degrades the quality factor and reduces the self-resonance frequency. The need of the hour is to obtain an inductor with high inductance density and high quality factor in the aforesaid frequency bands. Various attempts have been made to enhance the quality factor of integrated inductors on silicon by optimizing coil design, controlling substrate losses, using patterned ground shields, etc., with signifcant success. 5,6 However, it is understood that, to keep up with the technology scaling, on-chip inductors may not be practical to use beyond 10 GHz. 5 Distributed passive elements like transmission lines are suggested for the use at frequencies ≫10 GHz. But the length of such distributed elements can be prohibitive for their integration on-chip below 30 GHz. Integration of ferromagnetic and high-ε materials has thus become essential for frequencies of 10-30 GHz—some of the key frequency bands for 5G applications. Therefore, to continue with the miniaturization of RFICs and to move towards SoC solutions, on-chip integration of magnetic materials is necessary and that can pave way for high performance on-chip inductors for the frequency bands above 10 GHz, and thus can fulfll the promise of SoC and IoT. Magnetic materials have played a signifcant role in achieving effcient high frequency devices since the middle of the last century. Changing requirements over the years have sparked innovation in materials science and processing technology. At present, the requirement is to fig. 1. Proposed key performance indicators of upcoming 5G technology. (continued on next page) . © ECS 2018 ecsdl.org/site/terms_use address. See 207.241.231.83 Downloaded on 2018-07-21 to IP