IEEE Heterogeneous Integration Roadmap for 5G / 6G

Chapter 12 of the the IEEE  Heterogeneous Integration Roadmap covers the challenges for the 5G application space. It considers the needs of next generation of RF Front-End (RFFE) electronics to develop a roadmap for heterogeneously integrated components. It also considers the challenges for 6G (the longer 10 to 15 year timeframe).

It observes that in order to meet the ever more demanding use requirements within the radio system there will be a shift in design approach. The use of greater monolithic integration of systems function into more complex System on Chip designs will continue but an alternate approach, that looks for optimum solutions for radio functions using specific process fabrication technologies with heterogenous integration techniques,, will be adopted to bring the radio system together. For example in the Mobility Use Case for 5G, millimeter-wave signal propagation drives the need for smaller cells and may create demand for specific solutions for active, passive and radiators in 5G RF Front-Ends.

Due to allocation of different device operational frequency bands by different national spectrum license agencies designing a product that can act as a “world phone” will become more complex. In the 3G era only four bands were required. A 4G LTE phone needs to support more than 40 bands.  For 5G a world phone will likely add another 6 to 10 bands. This leads to increased complexity and needs much higher levels of component integration, another motivation for the role of heterogenous integration. There is discussion of the beamforming technologies that will require changes to the different radio architectures for the base-station and the end User Equipment (UE).

Three radio architecture challenges areas are identified: 

1. small element-to-element spacing needed for the phased-array  
2. appropriate semiconductor technology based on output power  
3. Front-End Modules integration level and packaging

There is a view that space is becoming an issue in the assembly of discrete components onto a PCB with an attached separate antenna. Active semiconductor devices are needed to achieve the required gain, output power, linearity and noise figure to achieve a high-fidelity communication links. The Front-End Modules might be best served by modules for added radio bands. These design challenges should lead to a range of radio architectures and HI packaging solutions to meet the range of market needs. 

There are sections covering:

• brief discussion on the inter-disciplinary challenges associated technologies, design techniques, simulation tools, materials and processes
• current trends
• tradeoffs between Antenna in Package (AiP) and Antenna on Chip (AoP) solutions

In terms of current trends, it is feasible for 5G millimeter-wave FEMs operating between 25 and 40 GHz to continue to take advantage of monolithic integration of functions enabled by CMOS and Fully-Depleted Silicon on Insulator (FD-SOI) to put analog, RF and digital IP blocks into a System on a Chip (SOC) for the highest volume applications. Increasing digital signal path density will drive down linewidth and spacing requirements. 

For heterogenous integration there should be opportunity for glass substrates since glass offers the lowest insertion loss and has excellent surface roughness. However, glass and other substrates have Coefficient of Thermal Expansion mismatches compared to the silicon based die and packaging technologies must address connection reliability as well as thermal and mechanical properties. There is useful information on materials and properties of substrates with characteristics that are relevant at 5G millimeter-wave bands. There is also discussions on the comparison of die based GaAs-based filters with PCB-based filters.

The relative advantages and disadvantages of on chip or in package antenna are discussed along with the issues of matching bandwidth and radiation efficiency to the permittivity / resistivity of the materials and substrates. As advances occur in heterogenous integration of components there will be issues of mechanical stability, manufacturing processes, system and component lifetimes, and increasingly, sustainability and circular economy considerations as part of new designs.

The final section 6 looks at 6G – Beyond 10 years where RFICs with large numbers of antenna elements will be needed, different modulation schemes, different material substrates for RFIC power amplifiers, increasing interconnect densities and tighter integration between the IC components and the antenna(s). Antenna-in-Package will play a major role for RF Front-End Modules that operate up to 60 GHz. Above 60 GHz, Antenna-on-Chip may become an attractive method for hardware integration.